(Go to Top Page)

The 31st International Technical Conference on Circuits/Systems, Computers and Communications
Technical Program

Remark: The presenter of each paper is marked with "*".
Technical Program:   SIMPLE version   DETAILED version with abstract
Author Index:   HERE

Session Schedule


Sunday, July 10, 2016

Registration (Foyer)
15:00 - 18:00
Welcome Reception (Hall)
18:00 - 20:00



Monday, July 11, 2016

FoyerHallRoom 1Room 2Room 3Room 4Room 5Room 6

M1-H  Mobile & Wireless Communications (1)
9:00 - 10:20
M1-1  [SS6] Network Virtualization Technologies
9:00 - 10:20
M1-2  VLSI Design
9:00 - 10:20
M1-3  [SS2] Advanced Data Storage Technology (1)
9:00 - 10:20
M1-4  [SS3] Electronics and Communication Technologies for Technical Education
9:00 - 10:20
M1-5  Image Processing (1)
9:00 - 10:20
M1-6  Computer Systems & Application (1)
9:00 - 10:20
Coffee Break (Foyer (2F))
10:20 - 10:40
Openining Ceremory (Hall)
10:40 - 11:00

PL1  Plenary Session I
11:00 - 12:00






Lunch
12:00 - 13:30
P1  Poster Session I
13:30 - 15:30
M2-H  Mobile & Wireless Communications (2)
13:30 - 15:30
M2-1  [SS1] Mathematical Systems Science and its Applications (1)
13:30 - 15:30
M2-2  VLSI Design & Testing
13:30 - 15:30
M2-3  [SS2] Advanced Data Storage Technology (2)
13:30 - 15:30
M2-4  [SS5] Video Signal Processing Algorithms and Systems
13:30 - 15:30
M2-5  Image Processing (2)
13:30 - 15:30
M2-6  Computer Systems & Application (2)
13:30 - 15:30
Coffee Break (Foyer (2F) and Room 7 (4F))
15:30 - 15:50
P2  Poster Session II
15:50 - 17:50
M3-H  Mobile & Wireless Communications (3)
15:50 - 17:50
M3-1  [SS1] Mathematical Systems Science and its Applications (2)
15:50 - 17:50
M3-2  Verification and Testing
15:50 - 17:50
M3-3  RF Circuits / Semiconductor Devices and Technology
15:50 - 17:50
M3-4  [SS4] Technologies for Beyond 4G Wireless Communications/Networks
15:50 - 17:50
M3-5  Image Processing (3)
15:50 - 17:50
M3-6  Computer Systems & Application (3)
15:50 - 17:50



Tuesday, July 12, 2016

FoyerHallRoom 1Room 2Room 3Room 4Room 5Room 6


T1-1  Mobile & Wireless Communications (4)
9:00 - 10:20
T1-2  Analog Circuits (1)
9:00 - 10:20
T1-3  Graphs and Net Theory
9:00 - 10:20
T1-4  Computers
9:00 - 10:20
T1-5  Computer Vision (1)
9:00 - 10:20

Coffee Break (Foyer (2F))
10:20 - 10:30

PL2  Plenary Session II
10:30 - 11:15







PL3  Plenary Session III
11:15 - 12:00






Lunch
12:00 - 13:30
P3  Poster Session III
13:30 - 15:30

T2-1  Mobile & Wireless Communications (5)
13:30 - 15:30
T2-2  Analog Circuits (2)
13:30 - 15:30
T2-3  [SS7] Information and Communication Technologies for Safe and Secure Life
13:30 - 15:30
T2-4  Internet Technology & Applications
13:30 - 15:30
T2-5  Computer Vision (2)
13:30 - 15:30
T2-6  Image Coding & Analysis
13:30 - 15:30
Coffee Break (Foyer (2F) and Room 7 (4F))
15:30 - 15:50


T3-1  [Tutorial] Unipolar multicarrier modulation for optical wireless communications with intensity modulation and direct detection
15:50 - 17:50
T3-2  Analog Circuits (3) and CAD
15:50 - 17:50
T3-3  Power Electronics & Control
15:50 - 17:50
T3-4  Security
15:50 - 17:50
T3-5  Communications (1)
15:50 - 17:50
T3-6  Multimedia Information Processing
15:50 - 17:50
Banquet (Hall)
18:30 - 21:00



Wednesday, July 13, 2016

FoyerHallRoom 1Room 2Room 3Room 4Room 5Room 6


W1-1  Artificial Intelligence (1)
9:00 - 10:20
W1-2  Intelligent Transportation Systems
9:00 - 10:20
W1-3  Antenna & Wave Propagation
9:00 - 10:20
W1-4  Optical Communications and Components (1)
9:00 - 10:20
W1-5  Communications (2)
9:00 - 10:20
W1-6  Audio / Speech Signal Processing (1)
9:00 - 10:20
Coffee Break (Room 7)
10:20 - 10:30


W2-1  Artificial Intelligence (2)
10:30 - 11:50
W2-2  Medical Electronics & Circuits
10:30 - 11:50

W2-4  Optical Communications and Components (2)
10:30 - 11:50

W2-6  Audio / Speech Signal Processing (2)
10:30 - 11:50


List of papers

Remark: The presenter of each paper is marked with "*".

Monday, July 11, 2016

Session PL1  Plenary Session I
Time: 11:00 - 12:00 Monday, July 11, 2016
Location: Hall
Chair: Tomohisa Wada (University of the Ryukyus, Japan)

PL1-1 (Time: 11:00 - 11:45)
TitleDependable Wireless BAN of Things beyond IoT - Its R&D, International Standard IEEE802.15, Global Business
AuthorRyuji Kohno (Yokohama National University, Japan)
KeywordBAN, Wireless Body Area Network, IEEE802.15
AbstractWireless body area network (BAN) has been researched and developed for ubiquitous and remote medicine and its international standard IEEE802.15.6 was established in February, 2012. Highly reliable and secure, i.e. dependable BAN can be applicable to a body of cars, buildings as well as a human body for dependable machine to machine (M2M) sensing and controlling. Such a M2M network can be called as "BAN of Things" like Internet of Things (IoT). To perform dependability of BAN, space and time signal processing technologies in physical layer must be jointly optimized with MAC, Network, and application layers. Even after BAN has been developed and standardized in global, regulatory science must be keen to guarantee safety, reliability and security to be compliant for regulation. This talk will introduce concept and possible manners of dependability in wireless BAN for medical healthcare, energy follow control, car controlling harness etc, i.e. Dependable BAN of Things. The speaker has been chairing a new standard activity for dependable wireless IEEE802.15 IG-DEP including amendment of BAN standard IEEE802.15.6.



Tuesday, July 12, 2016

Session PL2  Plenary Session II
Time: 10:30 - 11:15 Tuesday, July 12, 2016
Location: Hall
Chair: Heung-No Lee (Gwangju Institute of Science and Technology, Republic of Korea)

PL2-1 (Time: 10:30 - 11:15)
TitleWireless Energy Transfer Techniques for Future Communication Systems
AuthorInkyu Lee (Korea University, Republic of Korea)
KeywordWireless Energy Transfer
AbstractIn recent years, energy harvesting (EH) communication systems, where communication nodes harvest renewable energy from nature such as solar or wind, have attracted great attentions owing to its capability for supplying additional power to conventional energy-constrained systems. However, scavenging energy from nature may not be practical in some situations, for instance, when there are no energy sources around the energy-demanding devices. In addition, for small wireless devices such as sensor nodes or mobile phones, it would be difficult to implement EH circuitries which gather energy from solar or wind because of limited spaces. To overcome these issues, radio-frequency (RF) signals have been considered as a new energy source for wireless devices. In EH systems relying on the RF signals, a transmitter sends the RF signals to energy-demanding devices, and the energy of the RF signals can be harvested at the device. In contrast to other EH systems, such RF signal based EH techniques enable to charge energy-demanding devices whenever it is necessary. Recent experimental works have verified the feasibility of the RF signal based EH systems. In conventional wireless communication systems, the role of the RF signals has been limited to conveying information, although the RF signal can carry both information and energy concurrently. In this talk, I will describe wireless energy transfer systems based on RF signals, namely Simultaneous wireless information and power transfer (SWIPT) and Wireless powered communication networks (WPCN). First, several wireless energy transfer techniques will be introduced which utilize communication links for conveying energy to end users. Then, I will present a new technology which allows energy transfer in a multi-cell environment. Finally I will discuss challenges in research on wireless energy transfer systems and examine the possibility of application in future communication systems.


Session PL3  Plenary Session III
Time: 11:15 - 12:00 Tuesday, July 12, 2016
Location: Hall
Chair: Piya Kovintavewat (Nakhon Pathom Rajabhat University, Thailand)

PL3-1 (Time: 11:15 - 12:00)
TitleNon-Destructive Measurement Algorithms and Systems for Healthy Food
AuthorKosin Chamnongthai (King Mongkut's University of Technology Thonburi (KMUTT), Thailand)
KeywordNon-Destructive Measurement
AbstractHuman bodies naturally require different types and quantities of minerals and vitamins. We sometimes have a question why we do not consume only necessary foodstuff in adequate amount and quality. In order to feed the needed minerals and vitamins to human body, first of all, we have to sense what body needs for growth, health maintenance, and disease treatment at the moment. The needed things should be then converted into natural materials for cooking in the menu that a consumer prefers. The cooked food should be finally guaranteed in term of calories, needed minerals and vitamins, and medicine before consuming. In order to create this kind of healthy food production, human body has to be always monitored, and needed minerals and vitamins should be calculated. The foodstuffs including needed minerals and vitamins, which will be associated to be a kind of food in the menu that consumer wants, should be recommended by food-knowledge expert system. The cooked food also has to be measured in term of calories and important components. To meet this scenario, we have developed some algorithms to monitor and sense human body condition such as monitoring algorithm of blood vessels in retina, noninvasive-femur-bone-volume estimation algorithm for osteoporosis diagnosis, tonsillitis-monitoring algorithm, and breast-cancer monitoring algorithm. The systems for non-destructive measurement of fruit maturity, and measurement of moisture in granular material also have been developed as samples for auto-inspection of foodstuffs, and an algorithm for calories checking of food has been introduced and discussed in order to guarantee as healthy food.



Monday, July 11, 2016

Session M1-H  Mobile & Wireless Communications (1)
Time: 9:00 - 10:20 Monday, July 11, 2016
Location: Hall
Chair: Heung-Gyoon Ryu (Chungbuk National University, Republic of Korea)

M1-H-1 (Time: 9:00 - 9:20)
TitleNonlinearity Sensitivity of FBMC and UFMC Modulation System for the 5G Mobile System
Author*Changyoung An, Heung-Gyoon Ryu (Chungbuk National University, Republic of Korea), Yukinobu Fukushima, Tokumi Yokohira (Okayama University, Japan)
Pagepp. 1 - 4
Keyword5G, FBMC, UFMC, OFDM, nonlinearity
AbstractRecently, many researches on new waveforms have been carried out. UFMC and FBMC technique are famous as a new waveform for 5G. Similar to OFDM system, these techniques use multi-carrier. So, these systems have high PAPR, which can cause nonlinear distortion and OOB power increase because of the nonlinear HPA. In this paper, we have focused on spectrum characteristic analysis and BER performance evaluation of FBMC and UFMC system under the effect of nonlinear HPA. As simulation results, it is confirmed that OOB power of each system increases, and OOB power increase of FBMC system is the biggest. Additionally, we have confirmed that performance of every system is degraded by strength of HPA nonlinearity, and every system needs the PAPR reduction method for the nonlinear distortion compensation and power saving, even though it would be more complicated.

M1-H-2 (Time: 9:20 - 9:40)
TitleImproved Performance of 3G1X CDMA2000 Wireless Packet Data System in a Simulated Field Environment
Author*Kumar Venkata (HCL Technologies Limited, India)
Pagepp. 5 - 8
KeywordMS, BTS, PPP, PCF, PDSN
AbstractThe Third Generation (3G) CDMA2000 wireless telecommunication network standard has been designed to provide high speed data services on both forward and reverse link traffic's. In addition to the increased voice capacity, system performances and high-speed packet data services, it also supports quality of service for various end-user applications. However, the forward packet data link in 3G systems are time multiplexed among all of the mobile users. This paper addresses both the forward and reverse link packet data throughput performances in a simulated Additive white Gaussian noise (AWGN) environment and compare the results with the benchmarked 3G field data.

M1-H-3 (Time: 9:40 - 10:00)
TitleStudy on using Design Patterns to Implement a Simulation System for WiMAX Network
Author*Hsin-Hung Hsieh, Bih-Hwang Lee (National Taiwan University of Science and Technology, Taiwan), Huai-Kuei Wu (Oriental Institute of Technology, Taiwan), Wen-Pin Hsu, Hung-Chi Chien (National Taiwan University of Science and Technology, Taiwan)
Pagepp. 9 - 12
Keywordsimulator, design pattern, IEEE 802.16, object oriented design
AbstractA network simulator always is an important tool to observe and evaluate the study concept for wireless networks. If a selected simulator inherently has poor architecture, the maintenance, recondition, and expansion of functions will become more difficult and more time-consuming in the future, while it is difficult to understand and reuse by the successors. In order to make a simulator have flexible architecture and believable results, design patterns are proposed as the norms to design system architecture. A network simulator architecture is proposed, named as CCGns, which is a discrete-event virtual network simulator and follows the IEEE 802.16-2009 standard. CCGns obeys the object oriented design principles and is coded by the Java language. The main contribution includes three aspects which propose a scalable MAC messages management and the corresponding architecture, an applicable for multi-hop relay network topology architecture, and a two-stage minimum variance bandwidth allocation algorithm. By using mathematic calculation to verify the simulation results, the proposed system architecture has been proven to possess excellent fidelity.

M1-H-4 (Time: 10:00 - 10:20)
TitleAn Enhanced Autonomous Component Carrier Selection Method for HeNB in LTE-A Networks
Author*Bih-Hwang Lee, Hsin-Hung Hsieh (National Taiwan University of Science and Technology, Taiwan), Huai-Kuei Wu (Oriental Institute of Technology, Taiwan), Hung-Chi Chien, Wen-Pin Hsu (National Taiwan University of Science and Technology, Taiwan)
Pagepp. 13 - 16
Keyword3GPP, LTE-A, heterogeneous networks, carrier aggregation
AbstractBased on the IMT-Advanced standard, 3GPP defines the next-generation wireless network standard known as Long Term Evolution-Advanced. It provides relevant techniques to maximize spectrum efficiency, including heterogeneous networks and carrier aggregation. Carrier aggregation technique make it possible that the bandwidth can reach 1 Gbps when stationary and up to 100 Mbps when high speed mobility. However, it is an important issue to select primary component carrier and secondary component carrier to reduce interference and increase the number of user connections. We propose a component carrier reselection method based on autonomous component carrier selection but consider the uncoordinated interference. According to simulation results, the proposed method has better performance in fairness, carrier utilization and network system capacity.


Session M1-1  [SS6] Network Virtualization Technologies
Time: 9:00 - 10:20 Monday, July 11, 2016
Location: Room 1
Chair: Yoshihiro Kaneko (Gifu University, Japan)

M1-1-1 (Time: 9:00 - 9:20)
TitleContent Retrieval Method in Cooperation with CDN and Breadcrumbs Scoping on Domain
Author*Yutaro Inaba, Yosuke Tanigawa, Hideki Tode (Osaka Prefecture University, Japan)
Pagepp. 17 - 20
KeywordBreadcrumbs, Breadcrumbs+, cache, content-oriented-network
AbstractThese days, in addition to host-to-host communication, Information-Centric Network (ICN) has emerged to reflect current content-centric network usage. The authors aim to establish the feature of ICN on IP network to achieve feasible and efficient architecture. Specifically, we have proposed to operate Content Delivery Network (CDN) and Breadcrumbs (BC) frameworks coordinately on IP network. In this paper, furthermore, we propose to apply BC-Scoping on domain, which is a limited and selective BC framework as for network domain, to the cooperation.

M1-1-2 (Time: 9:20 - 9:40)
TitlePrototype Development of a Virtual Network Embedding System Using OpenStack
Author*Itsuho Goda, Kohei Sato, Yukinobu Fukushima (Okayama University, Japan), Heung-Gyoon Ryu (Chungbuk National University, Republic of Korea), Tokumi Yokohira (Okayama University, Japan)
Pagepp. 21 - 24
Keywordnetwork virtualization, virtual network embedding, OpenStack
AbstractNetwork virtualization enables us to make efficient use of resources in a physical network by embedding multiple virtual networks on the physical network. In this paper, we develop a prototype of a virtual network embedding system. Our system consists of OpenStack, which is an open source cloud service platform, and shell scripts. The shell scripts automatically construct the required virtual network on the physical network using OpenStack command-line interface. Experimental evaluation confirms that our system constructs a requested virtual network and assigns computing and networking resources to it appropriately.

M1-1-3 (Time: 9:40 - 10:00)
TitleDesign and Implementation of Proactive Firewall System in Cooperation with DNS and SDN
AuthorTomokazu Otsuka (Okayama University, Japan), *Nariyoshi Yamai (Tokyo University of Agriculture and Technology, Japan), Kiyohiko Okayama (Okayama University, Japan), Yong Jin (Tokyo Institute of Technology, Japan), Hiroya Ikarashi, Naoya Kitagawa (Tokyo University of Agriculture and Technology, Japan)
Pagepp. 25 - 28
KeywordFirewall, DNS, SDN, Traffic Engineering
AbstractRecently, unauthorized accesses from the external networks to the internal hosts are sharply increasing. Although many firewall appliances are widely utilized as one of the countermeasures, its throughput is not high enough especially when it performs deep packet inspection. In order to solve this problem, we propose a proactive firewall system which consists of two or more firewall appliances with Software Defined Network (SDN) adaptively choosing a proper one for each communication flow based on, for example, whether its peer is trusted or not. To obtain the peer IP address in advance, the system introduces EDNS Client Subnet option of DNS. According to the performance evaluation results on the prototype system, we confirmed that the prototype system could separate flows of trusted hosts from other flows effectively.

M1-1-4 (Time: 10:00 - 10:20)
TitleA Distributed Mobility Management Scheme in Flat Mobile Architecture
Author*Hua Yang, Naoki Wakamiya (Osaka University, Japan)
Pagepp. 29 - 32
Keyworddistributed mobility management, virtual servers, response threshold model, C-plane
AbstractTo accommodate exploding mobile data traffic, considerable number of devices, and heterogeneous applications a lot of researches on mobile core networks are oriented from hierarchical to flat architecture. In this paper, we first propose a novel flat architecture to separate the Cplane mobility management function to two virtual servers, i.e. VDMMEs for mobility management and DMDs for mobility databases. Then, we propose a mechanism and an algorithm based on the response threshold model to dynamically and adaptively allocate a VDMME and a DMD appropriate for an MN. We show our proposal can achieve lower delay and lower C-plane overhead than a comparative method.


Session M1-2  VLSI Design
Time: 9:00 - 10:20 Monday, July 11, 2016
Location: Room 2
Chair: Kimiyoshi Usami (Shibaura Institute of Technology, Japan)

M1-2-1 (Time: 9:00 - 9:20)
TitleDeadline-Constrained Static Mapping of Parallelizable Tasks on Manycore Architectures
Author*Yining Xu, Ittetsu Taniguchi, Hiroyuki Tomiyama (Ritsumeikan University, Japan)
Pagepp. 33 - 36
Keywordtask mapping, manycore SoCs, embedded systems
AbstractThis paper proposes a static task mapping technique for manycore architectures. The technique tries to minimize the number of cores while satisfying deadline constraints of individual tasks.

M1-2-2 (Time: 9:20 - 9:40)
TitleDesign and Evaluation of MTJ-based Standard Cell Memory
Author*Junya Akaike, Masaru Kudo, Kimiyoshi Usami (Shibaura Institute of Technology, Japan)
Pagepp. 37 - 40
Keywordpower-gating, MTJ, standard cell memory, break-even time
AbstractWith the spread of portable devices, products with high performance and long battery life are required. In this paper, we propose an MTJ-based SCM to reduce the power consumption of the cache memory. Simulation results showed that SCM can operate at a low voltage and has a high effectiveness in power reduction by PG.

M1-2-3 (Time: 9:40 - 10:00)
TitleLow-power Standard Cell Memory using Silicon-on-Thin-BOX (SOTB) and Body-bias Control
Author*Yusuke Yoshida, Masaru Kudo, Kimiyoshi Usami (Shibaura Institute of Technology, Japan)
Pagepp. 41 - 44
Keywordultra-low voltage operation, low-power, Standard Cell Memory, Silicon-on-Thin-BOX MOSFET, Body Bias
AbstractMemory which can reduce power consumption and realizes ultra-low voltage operation is required. This paper describes a design of low-power Standard Cell Memory (SCM) using Silicon-on-Thin-BOX (SOTB). Simulation results demonstrated that our SCM can reduce leakage current by 53% and energy consumption at the active mode by 70-85% as compared to the SRAM with the same circuit speed by body bias control. We also found that the SCM can operate at the voltage lower than 0.2V under process variation.


Session M1-3  [SS2] Advanced Data Storage Technology (1)
Time: 9:00 - 10:20 Monday, July 11, 2016
Location: Room 3
Chair: Rachsak Sakdanuphab (King Mongkut's Institute of Technology Ladkrabang, Thailand)

M1-3-1 (Time: 9:00 - 9:20)
TitleFabrication of Planar-Type Ti/TiOx/Ti Resistive Switching Device by SPM Local Oxidation
Author*Rachsak Sakdanuphab, Aparporn Sakulkalavek (King Mongkut’s Institute of Technology Ladkrabang, Thailand)
Pagepp. 45 - 48
KeywordRRAM, Scanning probe microscopy, Local anodic oxidation
AbstractMetal/Insulator/Metal (MIM) junctions are promising to be a resistive random access memory (RRAM). Almost MIM junction has been fabricated by stack-multilayer depositions. In this work, we study the planar-type Ti/TiOx/Ti junctions prepared by Ti sputtered film on SiO2/Si wafer, following with photolithography process, and local anodic oxidation by scanning probe microscopy (SPM). The growth and characterization of TiOx nanowires were investigated by atomic force microscopy (AFM). The oxide growth mechanism is proposed based on anodic oxidation theory. A resistive switching (RS) behavior of the junction was obtained by current-voltage measurement. Our results show the ability of SPM to growth TiOx nanowires as an insulator and the junctions have a resistive switching ability that can switch from a high-resistance state (HRS) to a low-resistance state (LRS) and from LRS to HRS.

M1-3-2 (Time: 9:20 - 9:40)
TitleSimplified Multi-track Joint 2-D Viterbi Detection for Bit Patterned Media Recording using Multi-head Array Technolgy
Author*Lin Min Min Myint, Thanatape Tubkaew (School of Information Technology, Shinawatra University, Thailand)
Pagepp. 49 - 52
KeywordViterbi algorithm, Bit pattern media reocording, 2-D Viterbi detection, multi-track multi-head, multi-track detection
AbstractVarious detection approaches were proposed to tackle 2-D interference in high density magnetic recording in the literature; however, most of them employed modified 1-D detectors, pseudo 2-D detectors extended from 1-D, instead of a full-fledged multi-track 2-D detection technique to avoid the high computational complexity. In this paper, firstly we study the performance of a full-fledged multi-track joint 2-D Viterbi detector exploiting multi-head array technology for bit patterned media recording (BPMR) system. Due to its high complexity, we also propose a simplified multi-track joint 2-D detector by doing a trade-off between complexity and performance gain. Finally, the performances of joint 2-D detectors are compared with that of conventional 2-D detection technique using multiple single-track detectors.

M1-3-3 (Time: 9:40 - 10:00)
TitleA Performance Improvement of a Rate-5/6 2D Modulation Code in Bit-Patterned Media Recording Systems
Author*Kanokleka Buahing, Wiparat Busyatras, Chanon Warisarn (College of Advanced Manufacturing Innovation, King Mongkut's Institute of Technology Ladkrabang, Thailand), Santi Koonkarnkhai, Piya Kovintavewat (Data Storage Technology Research Center, Nakhon Pathom Rajabhat University, Thailand)
Pagepp. 53 - 56
KeywordBit-pattern media recording, 2D interference, 2D Modulation code
AbstractA reduction in a track width in magnetic recording systems results in a delightful increase in areal density (AD), but also in an unpleasant appearance of intertrack interference (ITI). One way the effect of severe ITI may be mitigated is through the use of coding schemes. In this paper, we improve the performance of a rate 5/6 twodimension (2D) modulation code in bit-pattern media recording (BPMR) systems, and redesign code that efficiently eliminate ITI. Simulation results indicate that the proposed coding scheme is better than the conventional 2D modulation code, especially at high AD.


Session M1-4  [SS3] Electronics and Communication Technologies for Technical Education
Time: 9:00 - 10:20 Monday, July 11, 2016
Location: Room 4
Chair: Piya Kovintavewat (Nakhon Pathom Rajabhat University, Thailand)

M1-4-1 (Time: 9:00 - 9:20)
TitleDevelopment of Software defined Communication Systems for Improving Technical Education in Thailand
Author*Poompat Saengudomlert, Natthaphob Nimpitiwan, Pakorn Ubolkosold (Bangkok University, Thailand)
Pagepp. 57 - 60
Keyworddigital communications, laboratory, software defined
AbstractDue to high prices of imported laboratory equipment for teaching and learning digital communications, several educational institutes in Thailand (as well as some neighboring countries) are facing problems of equipment shortage. A lot of students end up learning from textbooks without actual experiments, causing them to lack confidence in their knowledge. Fortunately, reprogrammable digital electronics, in particular field programmable gate arrays (FPGAs), are becoming cheaper with higher computational powers. Universities can now use them to develop communication systems for laboratory equipment. This article describes such activities at Bangkok University-Center of Research in Optoelectronics, Communications, and Control Systems (BU-CROCCS), Thailand.

M1-4-2 (Time: 9:20 - 9:40)
TitleInteractive Browser-based Digital Communications Laboratory
AuthorPakorn Ubolkosold, *Poompat Saengudomlert, Natthaphob Nimpitiwan (Bangkok University, Thailand)
Pagepp. 61 - 64
Keyworddigital communications, signal processing, laboratory, web browser, interactive
AbstractBeing able to interact with and visualize the concept of digital communications through different forms of hands-on experiments is a crucial learning process for electrical/electronic engineering students to gain a better understanding of the subject. This paper presents a user-friendly browser-based digital communications training platform that allows students to interactively perform experiments covering topics like basic types of signals, lowpass filtering, matched filtering, baseband/passband modulations, signal multiplexing, and channel coding. Signals along the communication chain, i.e., at various points from the transmitter to the receiver, can be displayed and visualized on any standard web browser. Having a browser-based nature, the developed training platform allows for both offline and online uses. Consequently, it can be set to operate as a standalone, intranet-based, or Internet-based application.

M1-4-3 (Time: 9:40 - 10:00)
TitleWDM Planner: WDM Network Optimization Tool
AuthorPratkasem Vesarach (Independent Expert, Thailand), Wissarut Yutthachai (Electricity Generating Authority of Thailand, Thailand), *Poompat Saengudomlert (Bangkok University, Thailand)
Pagepp. 65 - 68
KeywordWDM network design, optimization, open-source software
AbstractWDM Planner is an open-source software for Wavelength Division Multiplexing (WDM) network optimization. It can determine the information such as the shortest paths between network nodes, the minimum amount of wavelength channels required to support the traffic demands, the maximum amount of traffic that the current network resources can support, and the minimum amount of additional wavelength channels required to support all traffic. It is a Graphical User Interface (GUI) based software. Accordingly, it can be conveniently used to help in actual planning of the WDM network resource allocation or to demonstrate the basic concepts in WDM network design for teaching and training purposes.

M1-4-4 (Time: 10:00 - 10:20)
TitleVisible Light Communication Development Kits Complianted to CP1223 Standard
AuthorPetch Nantivatana (Sripatum University, Thailand), Kata Jaruwongrungsee (National Electronic and Computer Technology Center (NECTEC), Thailand), Termpong Srited (Sripatum University, Thailand), *Piya Kovintaveewat (Nakhon Pathom Rajabhat University, Thailand), Preecha Kocharoen (Sripatum University, Thailand)
Pagepp. 69 - 72
Keywordvisible light communication, development kit
AbstractThe guideline of development platform both hardware and software for CP1223 standard has been introduced to accelerate the development of visible light communication (VLC) technology in Thailand. This development kit was developed by SARGMET researchers, which could be used to expedite the production time to market for industrial partners. The hardware is developed from the Arduino Pro micro (mini Leonardo) microcontroller, which is used to control a driver and a receiver circuit of the LED light source and the photo diode. The development kit also provides the microcontroller software scripts to create and control the I-4PPM modulation and serial data formatting according to the CP1223 standard defined. This can reduce the time to develop the VLC products with the ease of use and low complexity.


Session M1-5  Image Processing (1)
Time: 9:00 - 10:20 Monday, July 11, 2016
Location: Room 5
Chair: Katsuhiko Shimabukuro (University of the Ryukyus, Japan)

M1-5-1 (Time: 9:00 - 9:20)
TitleTime-consistent Depth Estimation for 2D-to-3D Conversion System Using Color Histogram and Variance Map
Author*Seungjun Shin, Kyeongbo Kong (POSTECH, Republic of Korea), Sung In Cho (LG Display Inc., Republic of Korea), Woo-Jin Song (POSTECH, Republic of Korea)
Pagepp. 73 - 76
KeywordSaliency, Depth-reversal, Estimation instability
AbstractWe present a time-consistent depth estimation algorithm for 2D-to-3D video conversion. The algorithm exploits variance of image to identify the foreground, and uses modified scene change detection method that exploits a color histogram. In simulations the proposed algorithm did not commit depth-reversal or saliency-instability.

M1-5-2 (Time: 9:20 - 9:40)
TitleHigh Level Synthesis of Neville Interpolation on an Embedded FPGA Platform using SDSoC
Author*Kyeong-Bin Park, Jung-Hyun Hong, Ki-Seok Chung (Hanyang University, Republic of Korea)
Pagepp. 77 - 78
KeywordInterpolation, High level synthesis, FPGA, Optimization
AbstractToday, high-level synthesis (HLS) has emerged as a widely used method for various digital systems. In this paper, we propose a the line interpolator implemented on Xilinx Zynq SoC using a high-level synthesis tool called Software-Defined SoC (SDSoC). There are lots of iterative calculations in the line interpolation algorithm; therefore, the hardware acceleration is preferred over the software-only implementation. We implement the Neville interpolation algorithm and apply the loop unrolling and pipelining techniques to fully utilize the target FPGA. When the utilization techniques are applied, the area is increased, but the total execution time is reduced by 10 times. The proposed implementation using SDSoC shows better performance compared to the software-only one, and fast design space exploration is achieved using HLS compared to the conventional RTL design.

M1-5-3 (Time: 9:40 - 10:00)
TitleAdaptive Multi-scale Self-similarity-based Enhancement Algorithm for Effective Up-scaling of Thermal Infrared Images
Author*Yong-Jun Kim, Byung Cheol Song (Inha University, Republic of Korea)
Pagepp. 79 - 80
KeywordIR image, enhancement
AbstractInfrared (IR) images usually have blurred edges and weak details in comparison with visible light images. Due to such phenomenon, up-scaled IR images do not provide acceptable visual quality, either. So, we propose an edge enhancement algorithm as a pre-processing for effective up-scaling of IR images. The proposed algorithm utilizes the block-based self-similarity which indicates local similarity between a current image and its scaled ones. First, for each block in the input IR image its edge strength is computed. Then, high-frequency information proper for the computed edge strength is derived. Here, the best self-example of the current block is found from a down-scale determined by the computed edge strength. By adding the extracted high-frequency information to the current block, the definition of the current block is enhanced. Experimental results show that the proposed algorithm provides better edge enhancement than state-of-the-art algorithms. Also, the proposed algorithm outperforms the other algorithm up to 0.07 even in terms of a quantitative metric, i.e., just noticeable blur metric (JNBM).


Session M1-6  Computer Systems & Application (1)
Time: 9:00 - 10:20 Monday, July 11, 2016
Location: Room 6
Chair: Seokhyun Yoon (Dankook University, Republic of Korea)

M1-6-1 (Time: 9:00 - 9:20)
TitleFPGA Implementation of Miller Line Encoding to Prevent Stuff Bits in CAN for Jitterless Communication
AuthorRonnie Opone Serfa Juan, *Min Woo Jeong, Hi Seok Kim (Cheongju University, Republic of Korea)
Pagepp. 81 - 84
KeywordMiller Line Encoding, bit stuffing, jitter, CAN, frame rate
AbstractController Area Network protocol utilizes Non Return-to-Zero (NRZ) line encoding, but long runs of consecutive bits with the same signal level may cause problems in transmission. Adding stuff bits can force the synchronization, however, it reduces the frame rate and causes jitter in communication. The main objective of this paper is to minimize or totally prevent the usage of stuff bits that causes jitter in CAN communication. Miller Line Encoding minimizes these drawbacks. The proposed scheme is synthesized on the Xilinx Virtex-5 FPGA. The results show that the Miller Line Encoding is better than the original line encoding.

M1-6-2 (Time: 9:20 - 9:40)
TitleLandslide-dammed Mapping and Logistic Regression Modeling Using GIS and R Statistical Software in The Northeast of Afghanistan
Author*Mohammad Kazem Naseri, Dongshik Kang (University of the Ryukyus, Japan)
Pagepp. 85 - 88
KeywordGIS and R, Logistic Regression, Landslide-dammed, Northeast of Afghanistan
AbstractIn this paper, we accomplished a logistic regression modeling to map and predict the probability of occurrences of lakes by landslides initiation. 361 lakes were mapped in the area. Simultaneously, 361 lakes point locations, 361 non-lakes location points were used for the model validation. Independent variables are mostly contributing for landslides lakes happening are a slope, distance to water sources, and relief. The result is a grid map that shows independent variables which are how statistically significant to the model.

M1-6-3 (Time: 9:40 - 10:00)
TitleProgrammable Logic Controller for the Process of Manufacturing Eyeglasses based on Program-Based Commissioning Method
Author*Pannawit Jungmanotham, Wudhichai Assawinchaichote (King Mongkut’s University of Technology Thonburi, Thailand)
Pagepp. 89 - 92
KeywordMobile Application, Programmable Logic Controller, Temperature Control
AbstractThe main contribution of this research is to develop an extra-verification for controlling the temperature during the production process by using Mobile Application together with Programmable Logic Controller based on Program-Based Commissioning Method. According to the simulated machine, the results have been found that this proposed method can minimize the possibilities of the product defects, and can also reduce the time-consuming production process. In addition, this proposed technique may be applied with other production processes as well.

M1-6-4 (Time: 10:00 - 10:20)
TitleEfficient Way-based Cache Partitioning for Low-Associativity Cache
Author*Byunghoon Lee, Eui-Young Chung (Yonsei University, Republic of Korea)
Pagepp. 93 - 96
Keywordcache management, shared last level cache, cache partitioning
AbstractCache Partitioning is well-known technique to reduce destructive interference among co-running applications in a shared last-level cache (SLLC). Way-based cache partitioning is a popular partitioning scheme due to its simplicity, but it can dramatically reduce associativity of each partition. Also, most SLLC have limited associativity because the higher associativity causes the higher cache access latency and power consumption. Therefore, we present Selective Cache Partitioning (SCP), a novel way-based cache partitioning technique for SLLC with the low associativity. SCP detects outstanding applications causing heavy cache pollution by on-line profiling and isolates them in private partitions. Then, it allocates non-outstanding applications to a shared partition, thereby partitioning SLLC selectively and providing more associativity to applications by sharing the partition. We provide experimental results to show the efficiency of SCP.


Session M2-H  Mobile & Wireless Communications (2)
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Hall
Chair: Inkyu Lee (Korea University, Republic of Korea)

M2-H-1 (Time: 13:30 - 13:50)
TitleRecognition of Wrist Position While Walking by Using Wearable Triaxial Accelerometers
Author*Kaoru Furumi, Shintaro Mizoguchi, Nanako Niioka, Masashi Imai, Atsushi Kurokawa (Hirosaki University, Japan)
Pagepp. 97 - 100
Keywordhuman activity recognition, wearable device, accelerometer
AbstractHuman activity recognition using wearable devices is becoming increasingly important. Techniques that estimate and recognize walking states while walkers carry an object in hand can be useful for safety and security. In this paper, we examine the recognition of walking states while walkers carry an object like a shoulder bag from data obtained by attaching a triaxial accelerometer to the wrists of walkers. The experimental results demonstrate that some states can be roughly identified from the accelerations and power spectrums.

M2-H-2 (Time: 13:50 - 14:10)
TitleThe Optimum Power Allocation Method in the Uplink of OFDMA System
Author*Zhang Wenheng, Asheralieva Alia, Yoshikazu Miyanaga (Hokkaido University, Japan)
Pagepp. 101 - 104
KeywordOFDMA, Power allocation, Wireless communication
AbstractThis paper proposes a novel adaptive transmission power and subcarrier allocation strategy for wireless communication system OFDMA (Orthogonal Frequency Division Multiple Access) that results in minimum power level transmission subject to target SNR constraints defined for each allocated subcarrier. And its advantages and disadvantages in comparison with traditional Waterfilling method.

M2-H-3 (Time: 14:10 - 14:30)
TitleComparison on Uplink Transmission Protocols for Two-User Wireless Powered Communication Network
Author*Yeonghwan Kim, Kuhyung Kwon, Chungyong Lee (Yonsei University, Republic of Korea)
Pagepp. 105 - 108
KeywordWPCN, DTS, Uplink protocol
AbstractLimited battery lifetime is a major consideration of wireless communication network systems. Wireless power transfer (WPT) is the emerging solution for the limited battery lifetime problem. Wireless powered communication network (WPCN) is one application of the WPT technology where devices are powered wirelessly by hybrid access point (H-AP) and transmit information to H-AP. In this paper, we adopt discrete time switching (DTS) protocol for two-user WPCN system. We propose two different uplink transmission protocols and present the performance comparisons of them.

M2-H-4 (Time: 14:30 - 14:50)
TitleAntenna Selection for Energy Efficiency in MIMO Wireless Energy Transfer Systems
Author*Sangwon Jung, Yonggue Han, Chungyong Lee (Yonsei University, Republic of Korea)
Pagepp. 109 - 112
KeywordEnergy transfer, Antenna selection
AbstractIn this paper, we consider MIMO wireless energy transfer system in which one energy receiver harvests energy from radio frequency signal transmitted from one energy transmitter. We deal with antenna selection problem to maximize harvested power to total power consumption ratio. It is shown that the singular value of the reconstructed channel by selection determines our objective. Against high computation complexity of exhaustive search for optimal selection, we proposed greedy selection algorithm based on the norm of column associated with each antenna. Simulation results show that the proposed algorithm achieves near optimal performance.

M2-H-5 (Time: 14:50 - 15:10)
TitleMatrix Decomposition Suitable for FPGA Implementation of NCSP-OFDM
AuthorHikaru Kawasaki, *Zhiheng Zhou, Masaya Ohta, Katsumi Yamashita (Osaka Prefecture University, Japan)
Pagepp. 113 - 116
KeywordOFDM, sidelobe suppression, precoding, FPGA implementation, singular value decomposition
AbstractN-continuous symbol padding orthogonal frequency division multiplexing (NCSP-OFDM) is a modulation technique that achieves a lower-sidelobe transmission signal by precoding data symbols. However, the precoding requires a wide dynamic range arithmetic. This study considers two decompositions of the precoder matrix for a low-end field-programmable gate array (FPGA) without DSP blocks. Results of numerical experiments and a test design confirm that the singular value decomposition is suitable for FPGA implementation.


Session M2-1  [SS1] Mathematical Systems Science and its Applications (1)
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Room 1
Chair: Qi-Wei Ge (Yamaguchi University, Japan)

M2-1-1 (Time: 13:30 - 13:50)
TitleEnumerating Minimal Siphons of Petri Net using SAT Solver
AuthorTan Lam Pham, *Atsushi Ohta, Kohkichi Tsuji (Aichi Prefectural University, Japan)
Pagepp. 117 - 120
KeywordPetri net
AbstractIn this paper, we study a method to enumerate all minimal siphons of Petri net. Conventional algorithm is based on Fourier-Motzkin method and if it fails because of memory shortage, no siphon is obtained. We suggest a successive enumeration algorithm using SAT solver. The solver can yield one solution to the logic equation for siphon. Then it is augmented clauses that exclude siphons that have been obtained. Computer experiment is done to show effectiveness of our method.

M2-1-2 (Time: 13:50 - 14:10)
TitleOn Each Condition of Soundness for Acyclic Free Choice Workflow Nets
Author*Shingo Yamaguchi, Naoki Nakahara (Yamaguchi University, Japan)
Pagepp. 121 - 124
KeywordPetri net, workflow net, soundness, liveness, boundedness
AbstractWorkflow nets (WF-nets for short) are Petri nets which represent workflows. Soundness is a criterion of logical correctness defined for WF-nets. A WF-net is said to be sound if it satisfies three conditions: (i) option to complete, (ii) proper completion, and (iii) no dead tasks. Our result shows that for an acyclic free choice WF-net, (1) Conditions (i) and (ii) of soundness are respectively equivalent to liveness and boundedness of the short-circuited net; (2) Checking of Conditions (i) and (ii) are respectively NP-complete; and (3) If the short-circuited net has no disjoint paths from a transition to a place (or no disjoint paths from a place to a transition), Conditions (i) and (ii) can be checked in polynomial time.

M2-1-3 (Time: 14:10 - 14:30)
TitleSupervisory Control of Concurrent Discrete Event Systems with Local Linear Temporal Logic Specifications
Author*Ami Sakakibara, Toshimitsu Ushio (Osaka University, Japan)
Pagepp. 125 - 128
Keywordsupervisory control, concurrent discrete event systems, linear temporal logic
AbstractWe consider a concurrent discrete event system consisting of N subsystems, where each subsystem has its local specification described by a linear temporal logic formula. We propose an algorithm to synthesize a supervisor for the concurrent system such that each subsystem satisfies a given linear temporal logic formula, and any subsystem never reaches a deadlock state.

M2-1-4 (Time: 14:30 - 14:50)
TitleToward Formal Analysis of Timed Anonymous Systems
Author*Yoshinobu Kawabe, Nobuhiro Ito (Aichi Institute of Technology, Japan)
Pagepp. 129 - 132
Keywordanonymity, timed systems, formal methods, verification, I/O-automaton
AbstractThis paper describes a basic idea to verify the anonymity of timed systems. Even though communication patterns are indistinguishable, the sender of a message can be identified by detecting the timing of message emission. In this paper we describe a timed system with an I/O-automaton-based formal specification language. By introducing a timer variable, we need to deal with an infinite-state system. With a simulation-based proof method for anonymity, we handle the infinite-state system directly.

M2-1-5 (Time: 14:50 - 15:10)
TitleAn MILP Approach to Optimal Surveillance over Graphs
Author*Mifuyu Kido, Koichi Kobayashi, Yuh Yamashita (Hokkaido University, Japan)
Pagepp. 133 - 136
Keywordoptimal surveillance, mixed integer linear programming
AbstractThe surveillance problem over graphs is to find an trajectory of an agent that travels each node as evenly as possible. This problem has several application such as city safety management and disaster rescue. In this paper, the finite-time optimal surveillance problem is formulated, and is reduced to a mixed integer linear programming (MILP) problem. Based on the policy of model predictive control, an optimal trajectory is generated by solving the MILP problem at each discrete time, and persistent surveillance can be realized.

M2-1-6 (Time: 15:10 - 15:30)
TitleEnergy-base Deadlock-free Supervisory Control of Quantitative Discrete Event Systems under Partial Observation
Author*Sasinee Pruekprasert, Toshimitsu Ushio (Graduate School of Engineering Science, Osaka University, Japan)
Pagepp. 137 - 140
KeywordSupervisory control, Discrete event system, Partial observation, Optimal control, Energy game
AbstractWe study the control of partially observed non-terminating quantitative DESs under the fixed-initial-credit energy objective. We model the control using a two-player game played between the supervisor and the DES on a weighted automaton. The DES aims to execute the events so that its energy level goes below zero, while the supervisor aims to maintain the energy level above zero. We show that the proposed problem is reducible to finding a winning strategy in a turn-based reachability game.


Session M2-2  VLSI Design & Testing
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Room 2
Chair: Tomohisa Wada (University of the Ryukyus, Japan)

M2-2-1 (Time: 13:30 - 13:50)
TitleHigh Performance and Low Design Cost TLB for MIPS32 Processor
Author*Takahiro Sasaki, Gun Muto, Yuki Fukazawa, Toshio Kondo (Mie University, Japan)
Pagepp. 141 - 144
KeywordProcessor architecture, TLB, RAM, CAM, MIPS32
AbstractTLB (Translation Lookaside Buffer) is one of the key components which affects performance and circuit scale. To improve performance of TLB, it is effective to increment the number of entries. However, because MIPS32 processor adopts CAM based TLB, increasing TLB entry causes serious enlargement of circuit scale. Furthermore, MIPS32 ISA limits the TLB entries up to 64. This paper proposes two methods to break the limitation; one is new approach to implement TLB using RAM for small footprint, and another is adopting a dynamic code analyzer to handle more than 64 TLB entries while keeping binary compatibility. According to the evaluation results, the proposed approach improves hit rate by 20% at the maximum, 6% in average, and reduces area by 29%.

M2-2-2 (Time: 13:50 - 14:10)
TitleA Design of 90 µW SENT Transmitter for Automotive Pressure Sensor in 0.18 µm CMOS Technology
Author*Imran Ali, Sung-Hun Cho, Dong-Soo Lee, Kang-Yoon Lee (Sungkyunkwan University, Republic of Korea)
Pagepp. 145 - 146
KeywordAutomotive Applications, Pressure Sensor, CMOS Technology, SENT, SAE J2716
AbstractIn this paper, a low power and area efficient SENT transmitter design for automotive pressure sensor is proposed. The performance analysis and simulation results shows that the design is fully compliant with SAE J2716 standard. The proposed design is implemented using 0.18 µm CMOS technology with a very small active area of 116 × 116 µm˛ and it requires only 4.314 K gates. The current consumption is 50 µA from a 1.8 V supply and it requires only 90 µW power. The design is also synthesized for different FPGA families from Xilinx and Altera. The minimum occupied slice registers are 52 and maximum operating frequency of 595 MHz is achieved. The proposed design is also tested with a standard SENT receiver module and 100 % accuracy with zero error is verified.

M2-2-3 (Time: 14:10 - 14:30)
TitleA Video Decoder Components Verification Scheme based on Transaction-level Information to Randomized Behavioral-level Operations Transformation
Author*Jiayi Zhu, Shinji Kimura (Waseda University, Japan)
Pagepp. 147 - 150
Keywordverification, video, decoder, transaction-level, behavioral-level
AbstractIn this article, we propose a video decoder components verification scheme based on the collaborations between reference software and VLSI test-bench. In the scheme, verilog models of DUT’s neighbors translate the transaction-level information generated by the reference software to behavioral-level operations which interact with the DUT component. In addition, timing of the behavioral-level operations of these verilog models are randomized to improve the verification efficiency.

M2-2-4 (Time: 14:30 - 14:50)
TitleComparative Analysis on Replica Techniques for Bit-Line Tracking in 14-nm node
Author*Se-Hyeok Oh, Han-Wool Jeong, Seong-Ook Jung (Yonsei University, Republic of Korea)
Pagepp. 151 - 154
Keywordreplica bit-line technique, TSAE, FinFET, delay, Energy consumption
AbstractReplica bit-line technique is used for generating sense amplifier enable signal which accurately tracks bit-line delay of SRAM. However, threshold voltage variation in the replica bit-line circuit changes the cell current, which results in variation of the sense amplifier enable time, TSAE. The variation of TSAE makes the sensing operation unstable. In this paper, in addition to conventional replica bit-line delay (RBLconv), dual replica bit-line delay (DRBD) and multi-stage dual replica bit-line delay (MDRBD) which are used for reducing TSAE variation are briefly introduced, and the maximum possible number of on-cell which can satisfy 6σ sensing yield is determined through simulation at a supply voltage of 0.6 V with 14 nm FinFET technology. As a result, it is observed that delay of DRBD and MDRBD are respectively improved by 24.4% and 48.3% compared with that of RBLconv and energy consumptions are respectively reduced by 8% and 32.4% compared with that of RBLconv.

M2-2-5 (Time: 14:50 - 15:10)
TitleEvaluation of Threshold Voltage Extraction Methods in Deep-submicron Technology
Author*Tae Hyun Kim, Hanwool Jeong, Seong-Ook Jung (Yonsei University, Republic of Korea)
Pagepp. 155 - 158
KeywordThreshold voltage, Deep-submicron, Extraction method
AbstractOld equations for threshold voltage aren’t suitable for short channel devices in thesedays. As an alternative of these equations, various medthods are devised to extract the threshold voltage from drain current versus gate voltage characteristic. This paper compares these methods with technology scaling in deep-submicrom technology and evluates which method is suitable

M2-2-6 (Time: 15:10 - 15:30)
TitleAutomatic Test Pattern Generation for Multiple Stuck-At Faults: When Testing for Single Faults is Insufficient
Author*Conrad JinYong Moore, Amir Masoud Gharehbaghi, Masahiro Fujita (University of Tokyo, Japan)
Pagepp. 159 - 162
KeywordAutomatic Test Pattern Generation, Multiple Stuck-at Fault
AbstractAs the number of gates per chip increase, the likelihood of fabricating a chip with multiple defects increases, yet there are no efficient ATPG techniques which can consistently detect such faults. The experiments and analysis in this paper show that, given an initial complete set of test patterns for single stuck-at faults, relatively few additional tests are necessary for full fault coverage. A potential method for covering all double stuck-at faults is briefly mentioned at the end of the paper.


Session M2-3  [SS2] Advanced Data Storage Technology (2)
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Room 3
Chair: Chanon Warisarn (King Mongkut's Institute of Technology Ladkrabang, Thailand)

M2-3-1 (Time: 13:30 - 13:50)
TitleEstimate Track Mis-registration Levels using an Energy Ratio Finding Technique in Bit-Patterned Media Recording Systems
Author*Wuttipong Ruksavong, Wiparat Busyatras, Chanon Warisarn (College of Advanced Manufacturing Innovation, King Mongkut's Institute of Technology Ladkrabang, Thailand), Lin M.M. Myint (School of Information Technology, Shinawatra University, Thailand), Piya Kovintavewat (Data Storage Technology Research Center, Nakhon Pathom Rajabhat University, Thailand)
Pagepp. 163 - 165
KeywordBit-patterned media recording(BPMR), Two-head two-track systems, Track mis-registration (TMR)
AbstractTrack mis-registration (TMR) is one of the undesirable effects in the ultra-high magnetic recording systems. Pragmatically, this effect has been handled from the servo system using the redundancy bits to identify the TMR levels, which requires some recording areas to store these redundancy bits. However, this paper proposes the energy ration finding technique based on the two readback signals to estimate the TMR levels in two-track two-head (2T2H) bit-patterned media recording (BPMR) systems. Then, the TMR effect is also corrected using the appropriate equalizers, which designed according to the estimated TMR levels. Simulation results indicate that the proposed technique can impressively estimate the TMR level and also yield the better performance than the conventional method especially when the TMR is large.

M2-3-2 (Time: 13:50 - 14:10)
TitleSoft Information Adjustment Scheme of a Rate-4/5 2D Modulation Code in Bit-Patterned Media Recording Systems
Author*Wongsatorn Teeraruangchaisri, Chavisa Kanjanakunchorn, Chanon Warisarn (College of Advanced Manufacturing Innovation, King Mongkut's Institute of Technology Ladkrabang, Thailand), Chaiyan Jettanasen (Faculty of Engineering, King Mongkut's Institute of Technology Ladkrabang, Thailand)
Pagepp. 167 - 170
KeywordBit-pattern media recording, Soft-output Viterbi algorithm, Two-dimensional interference
AbstractIn bit-pattern media recording (BPMR) systems, the readback signal of the desired track is severely interfered from the adjacent tracks, which known as the inter-track interference (ITI). One way to cope this problem is to use the 2-dimensional (2D) coding e.g., a rate-4/5 2D modulation code. Although the rate-4/5 modulation code ensures that the readback signal of the three inner tracks will not be corrupted by severe ITI; however, both the lowermost and the uppermost tracks can still be interfered by the outer tracks, which may lead to some errors in the data recovery process. To improve this shortcoming, we propose the soft-information adjustment scheme to adjust the soft-information before decoding. The relationship of the data encoding condition will be used as a criterion for adjusting the soft-information. Simulation results indicate that the proposed system is better than the conventional system, especially when AD and/or the position jitter are high.

M2-3-3 (Time: 14:10 - 14:30)
TitlePage-Based Weighted Bit-Flipping Decoding of Product LDPC Codes for Two-Dimensional Magnetic Recording
Author*Sirawit Khittiwitchayakul, Watid Phakphisut, Pornchai Supnithi (King Mongkut's Institute of Technology Ladkrabang, Thailand)
Pagepp. 171 - 174
KeywordTDMR, LDPC codes, Weighted bit-flipping
AbstractIn this work, the performance of product LDPC codes with page-based weighted bit-flipping (P-WBF) decoding is evaluated on the two-dimensional magnetic recording (TDMR) channel of which areal density approaching 5.62 Tb/in2. Furthermore, we propose to use multiple bits flipping to improve the decoding speed of P-WBF algorithm. The number of bits flipped can be a constant or inconstant at each iteration decoding. The simulation results show that the performance of P-WBF with multiple bits flipping is better than those of the P-WBF algorithm at small iteration number. However, its performance will suffer at high iteration.

M2-3-4 (Time: 14:30 - 14:50)
TitleBER Performance Study of Column Weight Two Non-Binary LDPC Codes with Predetermined Girth
Author*Nut Tantibut, Ambar Bajpai (Chulalongkorn University, Thailand), Kritsada Mamat (Kasetsart University, Thailand), Tharathorn Phromsa-ard (Chulalongkorn University, Thailand), Watid Phakphisut (Bangkok Suvarnabhumi University, Thailand), Piya Kovintavewat (Nakhon Pathom Rajabhat University, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 175 - 178
KeywordNon-Binary LDPC Codes, Column Weight Two, FFT
AbstractIn this paper, we constructed column weight two parity-check matrix by imposing pre-defined target girth and code length while maximizing code rate. Then this constructed binary parity-check matrix is transformed into non-binary matrix by using a randomly generated elements in Galios field GF(q), where q=2,4,8 and 16. We investigate bit-error rate performance (BER) for higher orders of Galios field GF(q) as compare to its binary counterpart. We found that for column weight two LDPC codes higher orders of GF(q) substantially improve BER performance and converge early than lower values of GF(q).


Session M2-4  [SS5] Video Signal Processing Algorithms and Systems
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Room 4
Chair: Hyuk-Jae Lee (Seoul National University, Republic of Korea)

M2-4-1 (Time: 13:30 - 13:50)
TitleTransform Domain Integer Distortion Estimation for HEVC Encoder
Author*Jeong-Mee Moon, Sungjei Kim, Jinwoo Jeong, Yong-Hwan Kim (Korea Electronics Technology Institute, Republic of Korea)
Pagepp. 179 - 181
KeywordRDO, mode decision, distortion, transform, quantization
AbstractHigh Efficiency Video Coding (HEVC) improved coding performance compared to earlier standard, H.264/AVC, while its computational complexity increased significantly. In order to find the optimal coding mode, Rate Distortion Optimization (RDO) searches all the possible modes to select the one having the minimum rate-distortion cost. The RDO framework shows best performance, but these holds most portion of the encoding complexity. This paper proposes the efficient way to find the optimal mode without a heavy computational burden of the RDO. We propose an integer level distortion estimation in transform domain and it can remove reconstruction processes needed for RDO mode decision. Experimental results show that the proposed distortion estimation can reduce encoding time about 13.9% with a negligible degradation of the Rate Distortion (RD) performance.

M2-4-2 (Time: 13:50 - 14:10)
TitleAdaptive Gaussian Filtering based on Contrast of Luminance for Perceptual Video Coding
Author*Wahyu Wiratama, Yong-Jo Ahn, Ismail Marzuki, Donggyu Sim (Kwangwoon University, Republic of Korea)
Pagepp. 183 - 184
Keywordpreprocessing, video coding, pre-filter, HEVC, Human Perceptual
AbstractThis paper proposes an adaptive Gaussian low pass filtering based on the contrast of luminance for perceptual video coding. The proposed adaptive Gaussian low pass filtering employs the contrast threshold of luminance to improve bit saving while maintaining the visual quality. In addition, a block-based filtering decision is also proposed to avoid the over smoothing effect around the high frequency area according to the edge information. Experimental results show that our proposed algorithm can save bitrate by 16% on average or up to 23% with trivially visual quality loss.

M2-4-3 (Time: 14:10 - 14:30)
TitleEfficient Schemes for Bypass Flag Read and Write for 1-D Lossless Frame Memory Compression in the Hardware-based Video Encoder
Author*Ji Hun Jang, Chae Eun Rhee (Inha University, Republic of Korea)
Pagepp. 185 - 188
Keywordbypass flag, frame memory compression, video codec, system bus, memory bandwidth
AbstractThe frame memory compression (FMC) is one of solutions to mitigate the memory size and bandwidth burdens for the reference frames used in the inter-prediction. The lossless FMC keeps the image quality, whereas the compression ratio is not guaranteed. In the worst case, the required memory size with FMC is more than that without FMC. To manage this case, the encoding step can be bypassed and the uncompressed data are stored with bypass flags. Mostly, the overhead from bypass flags depends on the size of compression units. However, if the bypass flags are stored to the external memory and writing and reading are done through the system bus, the trade-off between internal memory space and bus bandwidth needs to be considered carefully. In this paper, the schemes for bypass flag read and write are proposed to reduce the bandwidth effectively with a reasonable internal memory overhead.

M2-4-4 (Time: 14:30 - 14:50)
TitleParameter Decision for View Simulation in ASIFT Concerning the Affine Distortion Error
Author*Joohyuk Yum, Hyuk-Jae Lee (Seoul National University, Republic of Korea)
Pagepp. 189 - 192
KeywordASIFT, Affine transform, Error modeling
AbstractASIFT (affine invariant extension of SIFT) is fully affine invariant by simulating all image views obtained by varying viewpoints of the camera. This paper proposes an algorithm, which searches proper viewpoints for ASIFT algorithm in order to improve the robustness for the affine distortion. In this searching process, the affine distortion error proposed in this paper is used as the index of the distortion error caused by the viewpoint change. The ASIFT algorithm with the proposed viewpoints selection establishes correct matches by 1.18 times more than the conventional ASIFT algorithm, even though the computational complexity of the proposed method and the conventional method is same.

M2-4-5 (Time: 14:50 - 15:10)
TitleAn Efficient Hardware Architecture for Calculating the Minimum SAD with Parallel Execution at the Search Point Level
Author*Tae Sung Kim, Hyuk-Jae Lee (Seoul National University, Republic of Korea), Chae Eun Rhee (Inha University, Republic of Korea)
Pagepp. 193 - 196
KeywordHigh-Efficiency Video Coding, video compression, integer motion estimation, hardware architecture, Sum of absolute diffference
AbstractInteger Motion Estimation (IME) is one of the key components of video coding standard such as high-efficiency video coding (HEVC). As HEVC adopts a highly flexible block partitioning structure from 4x4 to 64x64, performing IME for every block partition demands considerable computational complexity. Although a number of previous works for efficient hardware architectures to accerarate IME have been done, the computational complexity of IME keeps growing as the video resolution increases. In this paper, an efficient hardware architecture for calculating the minimum SAD is proposed. The proposed hardware exploits the parallelism of multiple search points. To improve the efficiency of the proposed hardware, SAD calculation and its comparison steps are performed in a pipelined manner and the workload between the two pipeline stages are finely balanced. The proposed hardware processes 32 search points for an 8x8 block. The gate count is 827 K and the maximum operating clock frequency is 485.44 MHz.

M2-4-6 (Time: 15:10 - 15:30)
TitleMotion Estimation-Based Human Falling Detection for Visual Surveillance
Author*Heegwang Kim, Jinho Park, Hasil Park, Joonki Paik (Chung-Ang University, Republic of Korea)
Pagepp. 197 - 198
Keywordfalling detection
AbstractDetection of a human falling event has attracted increasing attention in a visual surveillance system. This paper presents a novel falling event detection algorithm using motion estimation and an integrated spatiotemporal energy map of the object region. The proposed method first extracts a human region using a background subtraction method. Next, we applied an optical flow algorithm to estimate motion vectors, and the energy map is generated by accumulating the detected human region for a certain period of time. We can then detect the falling event using the k-nearest neighbor (kNN) classification with the previously estimated motion information and energy map.


Session M2-5  Image Processing (2)
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Room 5
Chair: Koichi Ichige (Yokohama National University, Japan)

M2-5-1 (Time: 13:30 - 13:50)
TitleFast Image Compression Using Enhanced Singular Value Decomposition
Author*Harris Kristanto Husien, Wannida Sae-Tang (The Sirindhorn International Thai-German Graduate School of Engineering, King Mongkut's University of Technology North Bangkok, Thailand)
Pagepp. 199 - 202
KeywordExpected K-dimension, Singular value decomposition, Discrete cosine transform, JPEG, Image compression
AbstractThis paper proposes a new method called as expected K-dimension (EKD) for dealing with SVD image compression. The hardest thing is to find the best K-dimension or the redundancy level. The reason is that if the K-dimension that being chosen is too high, it means there is redundant data (more information can be cut), while if the K-dimension is set too low, it means some important information can be missing. The proposed method finds the minimum gap between values in S matrix and then utilizing the remaining values to determine the K-dimension. The results show that the proposed EKD method gives the result as good as the conventional SVD in terms of compression ratio and PSNR, but it reduces the processing time and the complexity drastically.

M2-5-2 (Time: 13:50 - 14:10)
TitleGranularly-Spatial Lossless Embedded Compression Algorithm on Memory Bandwidth Reduction for Video Codec Applications
AuthorGuo-Siang Huang, *Yu-Hsuan Lee (Yuan-Ze University, Taiwan)
Pagepp. 203 - 206
Keywordembedded compression(EC), granularly-spatial EC(GSEC), spatially adaptive Golomb-Rice code(SAGRC), spatially-finer prediction(SFP)
AbstractThe higher display resolution drives memory bandwidth issue to arise in video codec. The embedded compression (EC) is an efficient algorithm to reduce memory bandwidth. In this paper, a lossless EC using granularly-spatial EC (GSEC) algorithm is proposed. It comprises two core techniques of spatially-finer prediction (SFP) and spatiallyadaptive Golomb-Rice code (SAGRC). The SFP generates the residual for every pixel, and the SAGRC convert each residual to an efficient codeword. The gradient information around current pixel can make a link between SFP and SAGRC to improve coding efficiency. Experiment results reveal that this work can save memory bandwidth by 37% on average without any degradation on visual quality.

M2-5-3 (Time: 14:10 - 14:30)
TitleEvolution of Contours for Shape Recognition
Author*Ukrit Watchareeruetai, Kriangkrai Phanjan (King Mongkut’s Institute of Technology Ladkrabang, Thailand)
Pagepp. 207 - 210
Keywordshape, contour, feature extraction, linear genetic programming
AbstractThis paper proposes an evolutionary method for constructing contour feature extraction programs for shape recognition. The proposed method adopts a variant of genetic programming (GP), called linear GP, to optimize the performance of programs. Linear GP used in this paper requires two types of registers: 1) numerical registers and 2) contour registers. Consequently, the contour of an object which is stored in a contour register can be processed to produce some features and also be transformed by some primitive operators to generate another contour. During evolutionary process, an input contour is evolved and, hopefully, its useful features are then extracted. Preliminary results show that the proposed method can automatically construct a contour feature extractor for a leaf recognition problem, with an accuracy of 90%.

M2-5-4 (Time: 14:30 - 14:50)
TitleAn Adaptive Search Range Selection Algorithm for HEVC
Author*Yoshiki Ito, Wen Shi, Tian Song, Takashi Shimamoto (The University of Tokushima, Japan)
Pagepp. 211 - 214
KeywordMotion estimation, Search range, Complexity reduction, Coding efficiency, HEVC
AbstractIn this paper, we proposed an adaptive search range selection algorithm. In previous work, the method which determines the search range using motion vector variance was proposed. However, the selected search range is not efficient enough. The proposed algorithm can efficiently reduce the redundant search range using the distribution of the surrounding motion vectors. The simulation results show that the proposed algorithm achieved averagely 76.0\% complexity reduction with 0.4\% BD-rate increasing, compared with previous work.

M2-5-5 (Time: 14:50 - 15:10)
TitleColor Change Detection based on Wavelet Variance
Author*Jong Ju Jeon, Hyun Jun Shin, Il Kyu Eom (Pusan National University, Republic of Korea)
Pagepp. 215 - 216
Keywordwavelet transform, color modification, color change detection
AbstractColor modification is one of commonly used image forgery techniques. In this paper, we present a new estimation method of the color modification using variance ratio of color difference model and wavelet transform. The color difference model is used to emphasize the distortion in CFA pattern caused by color modification. We define variance ratio value to determine the modified hue angle. Experimental results show that the proposed method has good results in detecting color modification forgery. Compared with conventional method, our method provides reasonable outputs.

M2-5-6 (Time: 15:10 - 15:30)
TitleHardware Oriented Early CU Splitting Algorithm by Coding Unit Feature Analysis for HEVC
Author*Ryo Kuroda, Wen Shi, Tian Song, Takashi Shimamoto (The University of Tokushima, Japan)
Pagepp. 217 - 220
KeywordCoding Unit (CU), Coding Tree Unit (CTU), HEVC, Inter coding
AbstractIn this paper, a CU size early determination algorithm is proposed to efficiently reduce complexity for the hardware implementation of HEVC inter coding.The proposed algorithm can be positioned as a pre-processing for the rate-distortion optimization (RDO) which can decide the CU size by only analyzing the original pixels based on the characteristics analysis of the coding unit using sobel-filter.Simulation results show that the proposed algorithm will induce PSNR loss 0.042dB with only 1.72% bitrate increase, compared to HM9.1.


Session M2-6  Computer Systems & Application (2)
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Room 6
Chair: Takeo Okazaki (University of the Ryukyus, Japan)

M2-6-2 (Time: 13:30 - 13:50)
Title3D Neighborhood Relationship of Cellular Genetic Algorithm for Tour Guide Assignment Problem
Author*Lina Setiyani, Takeo Okazaki (University of the Ryukyus, Japan)
Pagepp. 225 - 227
Keyword3D Neighborhood, Tour Guide Assignment Problem, Cellular Genetic Algorithm
AbstractManagement optimization is very important in tourism, one of the problems in management is tour guide assignment. The objective of this research is to give a clear understanding into the advantages of increasing cellular dimensionality on Tour Guide Assignment Problem by using Cellular Genetic Algorithm.

M2-6-3 (Time: 13:50 - 14:10)
TitleWS model based Massively Parallel Genetic Algorithm and its Various Applications
Author*Kotaro Maekawa, Hajime Nobuhara (Department of Intelligent Interaction Technology, University of Tsukuba, Japan)
Pagepp. 229 - 232
KeywordGenetic Algorthm, WS model, Optimization
AbstractGenetic algorithms are widely used method as solutions for optimization problems. However in complex conditions, it does not work properly. In this paper, to solve this problem, we propose WS model GA based on Massively Parallel Genetic Algorithm, which has diversity. The system is applied to Function optimizations, and a classroom optimization, and it’s confirmed that WS model GA was better result than Standard GA and Massively Parallel Genetic Algorithm in complex experimental cases.

M2-6-4 (Time: 14:10 - 14:30)
TitleA Hardware Accelerator for Bag-of Features based Visual Word Transformation in Computer Aided Diagnosis for Colorectal Endoscopic Images
Author*Tetsushi Koide, Takumi Okamoto, Koki Sugi, Tatsuya Shimizu, Anh-Tuan Hoang, Toru Tamaki, Bisser Raytchev, Kazufumi Kaneda (Hiroshima University, Japan), Shigeto Yoshida, Hiroshi Mieno (Hiroshima General Hospital of West Japan Railway Company, Japan), Shinji Tanaka (Medicine Graduate School of Biomedical and Health Science, Hiroshima University, Japan)
Pagepp. 233 - 236
KeywordComputer-Aided Diagnosis, Endoscopic Image Processing, FPGA, Visual Word, Real-Time Processing
AbstractThis paper presents an FPGA based hardware accelarator for feature transformation in real-time computer-aided diagnosis (CAD) system for colorectal endoscopic images with narrow-band imaging (NBI) magnification. We have demonstrated the proposed FPGA implimentation was very compact, the permance met the crinical doctors needs (throuthput > 5 fps, latency < 1 sec), and it achiveded throughput: 16.7 fps and latency: 60 msec with keeping the quality of real-time diagnostic support.

M2-6-5 (Time: 14:30 - 14:50)
TitleParallel Application Performance Comparison with Vulkan, CUDA and OpenMP
Author*SeongHu Hong, DoHyeong Kim, Chang-Sung Jeong (Korea University, Republic of Korea)
Pagepp. 237 - 240
KeywordVULKAN, CUDA, Parallel Applications, Parallel Processing, GPGPU
AbstractVulkan is an API (Application Programming Interface) for graphics and compute hardware, launched by Khronos. For Programming model, it recently draws attention as a Graphic and Compute Integrated API, and much works have been conducted focused on Vulkan's Graphic API performance, but not so much works on Vulkan’s Compute API. Therefore, in this paper, we shall evaluate and compare the performance of Vulkan’s Compute API with respect to CUDA and OpenMP by implementing three parallel applications using them respectively. Also, we shall show that the performance of Vulkan is similar to that of CUDA, and has advantage of using graphic and compute operations at the same time.

M2-6-6 (Time: 14:50 - 15:10)
TitleConductivity Image Reconstruction Based on Singular Value Decomposition Method in Electrical Impedance Tomography
Author*Bong Seok Kim, Anil Kumar Khambampati, Suk In Kang, Kyung Youn Kim (Jeju National University, Republic of Korea)
Pagepp. 241 - 244
KeywordElectrical impedance tomography, Singular value decomposition, Tikhonov regularization, Image reconstruction
AbstractIn this paper, an inverse method based on singular value decomposition is proposed to solve the inverse problem with the generalized Tikhonov regularization to determine the internal conductivity distribution in electrical impedance tomography. Numerical simulations have been carried out to evaluate the performance of the proposed method.


Session M3-H  Mobile & Wireless Communications (3)
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Hall
Chair: Masato Saito (University of the Ryukyus, Japan)

M3-H-1 (Time: 15:50 - 16:10)
TitleRTL Design of FA-WLAN System and its Verification
Author*Mitsuru Hamada, Duc Khai Lam, Keishi Yamaguchi, Yuhei Nagao, Masayuki Kurosaki, Hiroshi Ochi (Kyushu Institute of Technology, Japan)
Pagepp. 245 - 248
KeywordFactory Automation, Industrial Wireless Local Area Network
AbstractIn this work, we propose an industrial Wireless Local Area Network(iWLAN) system for industrial robots(iRBs) control in factory automation(FA) environments. We apply a synchronous multi-user(MU) round-robin transmission protocol. The analysis and simulation results show that our proposals provide 100% better throughput than the conventional system. Furthermore, the architecture of iWLAN system is proposed and designed to successfully implement the proposed protocol. The RTL verification result of the proposed system is also shown.

M3-H-2 (Time: 16:10 - 16:30)
TitleComputational Complexity Reduction of Orthogonal Precoding of N-continuous OFDM
AuthorHikaru Kawasaki, *Takahiro Matsui, Masaya Ohta, Katsumi Yamashita (Osaka Prefecture University, Japan)
Pagepp. 249 - 252
KeywordOFDM, orthogonal precoding, singular-value decomposition, computational complexity reduction
AbstractN-continuous OFDM is a precoding method for sidelobe suppression of orthogonal frequency division multiplexing (OFDM) signals and is to seamlessly connect OFDM symbols up to the high order derivative for sidelobe suppression, which is suitable for suppressing out-of-band radiation. However, it degrade the error rate severely as increasing the continuous derivative order. Orthogonal precoding of N-continuous OFDM has both a sidelobe suppression performance and an ideal error rate; however, it requires a very large computational complexity for precoding and decoding. This paper proposes a matrix decomposition of the large-sized matrix in the orthogonal precoding of N-continuous OFDM to reduce the computational complexity. Numerical experiments show that the proposed method can drastically reduce the computational complexity.

M3-H-3 (Time: 16:30 - 16:50)
TitleSpectrum Characteristic Analysis of FBMC, UFMC, and W-OFDM System Under The Nonlinear HPA Environments
Author*Changyoung An, Kyunglok Kung, Byeongjae Kim, Heung-Gyoon Ryu (Chungbuk National University, Republic of Korea)
Pagepp. 253 - 256
Keyword5G, OOB, UFMC, FBMC, W-OFDM
AbstractIn order to satisfy requirements of 5G mobile communication, researches on new waveform based on multi-carrier are being carried out. In this paper, we describe OFDM, UFMC, FBMC, and W-OFDM. And then, we evaluate and analyze spectrum characteristics of these system under the nonlinear HPA environment. As simulation results, FBMC has the lowest OOB power under the linear HPA condition. But, OOB power levels of UFMC, FBMC, and W-OFDM are similar under the nonlinear HPA environment. In real implementation, HPA has nonlinear characteristic. Therefore, W-OFDM technique is more advantageous because W-OFDM system has low-complexity and requires simple processing

M3-H-4 (Time: 16:50 - 17:10)
TitlePAPR Reduction by Infinity-Norm Minimization in Blind PTS-OFDM Systems
Author*Moeko Yoshida, Teruyuki Miyajima (Ibaraki University, Japan)
Pagepp. 257 - 260
KeywordOFDM, PAPR, PTS
AbstractIn this paper, we propose a partial transmit sequences(PTS)-based PAPR reduction scheme without side information for OFDM systems. In the transmitter, an iterative algorithm which minimizes the infinity-norm of a transmitted signal is used to determine phase coefficients. To avoid the transmission of side information, the range of the phase coefficients is restricted. In the receiver, the phase coefficients are blindly estimated by evaluating the phase differences between equalizer outputs and candidate symbols. Simulation results show that the proposed PAPR reduction scheme is more computationally efficient than the conventional PTS. Moreover, the impact of the restricted phases on BER performance is evaluated.

M3-H-5 (Time: 17:10 - 17:30)
TitleAn Improved Framed slotted ALOHA-Based Anti-collision Algorithm with Skipping Idle Slots for RFID Systems
Author*Sanika Krishnamali Wijayasekara, Chairat Phongphanphanee (Department of Electrical Engineering Chulalongkorn University, Thailand), Robithoh Annur (Department of Computer and Communications Technology, Tunku Abdul Rahman University, Malaysia), Suvit Nakpeerayuth, Lunchakorn Wuttisittikulkij (Department of Electrical Engineering Chulalongkorn University, Thailand)
Pagepp. 261 - 263
KeywordALOHA
AbstractRecently a new framed slotted ALOHA based ant-icollision protocol has been proposed that yields better tag identification efficiency than all existing conventional identification algorithms. The protocol employs idle slot skipping method where the reader creates the subsequent frames by omitting the non selected slots, and uses the variable frame size of two or four. In this paper we investigate the effects of exploiting different frame sizes on the identification efficiency of the protocol. The simulation results shows that the performance of the original protocol can be enhance by fixing frame size to a larger number. The identification efficiency is significantly increased by 20% .

M3-H-6 (Time: 17:30 - 17:50)
TitleMulti-Camera Based Object-Retrieving by Integrating Electronic Signals
Author*Tingsong Chen, Cheng Zhang (Waseda University, Japan), Kyoko Yamori (Asahi University, Japan), Yoshiaki Tanaka (Waseda University, Japan)
Pagepp. 265 - 268
Keywordelectronic, visual, signal fusion, localization, retrieval
AbstractIn video surveillance system, video cameras continually generate a huge amount of data. Nowadays electronic signals, such as Wi-Fi, RFID, are pervasive, and each of them has a unique number which can be taken as the identifier of the target object. In this paper, a new technology is proposed for object-retrieve from video data by integrating electronic signal of the object. We use the best match of electronic signal and visual signal to localize the position of object more accurately.


Session M3-1  [SS1] Mathematical Systems Science and its Applications (2)
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Room 1
Chair: Shingo Yamaguchi (Yamaguchi University, Japan)

M3-1-1 (Time: 15:50 - 16:10)
TitleA Study on Synthesizing State Machines from Multiple Communication Diagrams
Author*Toshiki Kinoshita, Toshiyuki Miyamoto (Osaka University, Japan)
Pagepp. 269 - 272
Keywordunified modeling language, choreography realization problem, automatic synthesis, service-oreinted architecture, model-based development
AbstractFor a service-oriented architecture based system, the problem of synthesizing a concrete model for each service configuring the system from an abstract specification, which is referred to as choreography, is known as the choreography realization problem. In this paper, we assume that choreography is given by two acyclic relations. We study the conditions for acyclic relations to synthesize concrete models.

M3-1-2 (Time: 16:10 - 16:30)
TitleA Petri Net Model of Internal Organs and Pericardium Meridian Based on the Acupuncture Efficacy
Author*Masafumi Nishida (Yamaguchi University, Japan), Ren Wu (Yamaguchi Junior College, Japan), Mitsuru Nakata, Qi-Wei Ge (Yamaguchi University, Japan)
Pagepp. 273 - 276
KeywordTraditional Chinese Medicine, Modeling, Petri Net, Five Viscera and Six Bowels, Pericardium Meridian
AbstractThis paper aims at constructing a Petri net model of internal organs with triple energizer and pericardium meridian, and its simulation model based on Traditional Chinese Medicine (TCM). Till now, we have proposed a model of the internal organs without triple energizer. In this paper, we add triple energizer to propose a new model of the internal organs as well as its combined model with pericardium meridian according to the acupuncture efficacy. Then, we develop a related simulation model in order to automatically simulate the behaviour of the internal organs. Finally, we show simulation results by stimulating acupuncture points of the meridian.

M3-1-3 (Time: 16:30 - 16:50)
TitleComputational Models for the Design of Artificial Genetic Circuits of RS- and JK-FlipFlops
Author*Manabu Sugii, Masashi Kubota, Hiroshi Matsuno (Yamaguchi University, Japan)
Pagepp. 277 - 280
Keywordartificial genetic circuits, binary network, ordinary differential equation
AbstractSeveral kind of artificial genetic circuits (AGCs) have been designed with the knowledge of biology, mathematics and informatics. We analyzed dynamic models of the AGCs to realize RS- and JK-flipflops. We visualized nullcline diagrams of the ODEs for these flipflops in a phase space and confirmed the behavior of these dynamic models whether these models can function as designed. We also tried to design gene structures so that the regulation of gene expression is similar to these flipflop circuit functions.

M3-1-4 (Time: 16:50 - 17:10)
TitleImplementation and Evaluation of Similar Subgraph Retrieving
Author*Hiroaki Kodama, Mitsuru Nakata, Qi-Wei Ge, Makoto Yoshimura (Yamaguchi University, Japan)
Pagepp. 281 - 284
Keywordsimilar subgraph retrieving, Isomorphic subgraph, image retrieval for historical documents, feature graph
AbstractOur aim is to realize an image search system for Japanese historical document which doesn't require reprinting and revising processes. In our system, the structural information of a string written in historical documents is expressed by a feature graph. So far, we have realized a method of generating feature graphs and proposed an algorithm for similar subgraph retrieving. In this paper, we describe the implementation and evaluation of the similar subgraph search program.

M3-1-5 (Time: 17:10 - 17:30)
TitleOn Computation of Association Coefficients of Phonetic Symbols Based on Basic Vocabulary
Author*Ren Wu (Yamaguchi Junior College, Japan), Yuya Matsuura, Hiroshi Matsuno (Yamaguchi University, Japan)
Pagepp. 285 - 286
Keywordbasic vocabulary, phonetic symbols, phi coefficient, Levenshtein distance
AbstractIn the method of employing Levenshtein distance for computing linguistic distance, the biggest disadvantage is that, basic vocabulary is simply thought of pure strings that do not have any sound features. This paper reports a method computing the phi coefficient that is one kind of association coefficients of phonetic symbols used as weight coefficients of costs of the edit operations.

M3-1-6 (Time: 17:30 - 17:50)
TitleAutomated Generation of Mixed Integer Programming for Job-shop Scheduling Problems based on Petri Nets
AuthorRyosuke Ushijima, *Andrea Veronica Porco (University of the Ryukyus, Japan), Hideki Kinjo (Okinawa University, Japan), Morikazu Nakamura (University of the Ryukyus, Japan)
Pagepp. 287 - 290
KeywordPetri nets, Mathematical Programming, Scheduling Problem
AbstractThis paper proposes a scheme for automated generation of a mixed-integer programming for scheduling problems based on timed Petri nets. Our tool reads XML-based Petri net data, extract precedence and conflict relations among transitions and then generates a mixed integer programming for the target scheduling problem. Therefore, once users model their system with Petri nets, they can address the scheduling problem for efficient operations. For implementation, we use CPN Tools, a well-known tool for editing, simulating and analyzing Colored Petri nets, for modeling. Users can model their systems with CPN Tools and then generate a mixed-integer programming problem for scheduling.


Session M3-2  Verification and Testing
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Room 2
Chairs: Hideki Asai (Shizuoka University, Japan), Hiroyuki Yotsuyanagi (Tokushima University, Japan)

M3-2-1 (Time: 15:50 - 16:10)
TitleA Built-in Test Circuit for Injected Charge Tests of Open Defects in CMOS ICs
Author*Kouhei Ohtani, Daisuke Suga, Hiroyuki Yotsuyanagi, Masaki Hashizume (Tokushima University, Japan)
Pagepp. 291 - 294
Keywordtest, open defect, BIC, IC
AbstractIn this paper, a built-in test circuit for an electrical test method is proposed to detect open defects in CMOS ICs. The test method is based on amount of charge injected from a power supply voltage source. A memory IC is prototyped in which the test circuit is embedded. It is examined experimentally whether an open defect in the IC can be detected with the test circuit. The experimental results show that an open defect in ICs can be detected by the test method.

M3-2-2 (Time: 16:10 - 16:30)
TitleA Built-in Test Circuit to Monitor Changing Process of Resistive Open Defects in 3D ICs
Author*Masashi Okamoto, Akihiro Odoriba, Hiroyuki Yotsuyanagi, Masaki Hashizume (Tokushima University, Japan), Shyue-Kung Lu (National Taiwan University of Science and Technology, Taiwan)
Pagepp. 295 - 298
KeywordBIST, open defect, 3D IC
AbstractA resistive open defect in a 3D IC may change into a hard open one after shipping to a market. In this paper, a built-in test circuit is proposed to monitor the changing process of resisitive open defects occurring at interconnects between dies including an IEEE 1149.1 test circuit. Feasibility of the process monitor with the test circuit is examined by Spice simulation The simulation results show that the process of a resistive open defect can be monitored at a test speed of 1MHz per an interconnect.

M3-2-3 (Time: 16:30 - 16:50)
TitleRandom Forest Model for Silicon-to-SPICE Gap and FinFET Design Attributes Identification
AuthorHyosig Won, *Katsuhiro Shimazu (Samsung Electronics, Republic of Korea)
Pagepp. 299 - 302
Keywordrandom forest, attribute, finfet, spice silicon gap, vehicle
AbstractRandom forest model was applied to analyze design attributes influence on Silicon-to-SPICE(S2S) gap. In order to have enough model accuracy to discuss S2S gap, scaled learning data was used with random design attributes count to build each tree in forest model. From the improved model, indices so-called ‘importance’ and newly defined ‘impact’ can be extracted to identify significant design attributes which determine S2S gap. The identified design attributes classify S2S gap well and show clear trend of it. Finally the key FinFET structures can be identified as the representative layout structure to cause large S2S gap.

M3-2-4 (Time: 16:50 - 17:10)
TitleMoisture-Insensitive Low-Concentration Oxidizing-Gas Sensor
Author*Taku Horiguchi (Graduate School of Science and Technology, Nihon University, Japan), Yoshiki Sasaki, Katsutoshi Saeki (Colleges of Science and Technology, Nihon University, Japan)
Pagepp. 303 - 306
KeywordGas sensor, Moistureproof, Oxidizing gas, Si substrate, Complex impedance
AbstractWe propose a new sensor which is comb-shaped electrode on a Si substrate for moistureproof. We suggest a new measurement method for gas distinguish between NO2 and SO2 at low concentrations. We showed that the low concentration oxidizing gas sensor using a Si substrate has moistureproof within 60% and 80%. Furthermore, it is possible to distinguish from the difference between the change in the pattern of the complex impedance of NO2 and SO2 by using complex impedance method.

M3-2-5 (Time: 17:10 - 17:30)
TitleFPGA-Based Educational System Cooperating with Mobile Applications for Learning Computer Architecture
Author*Hiroyuki Maeda, Yoshihiro Yasutake, Azumi Iihoshi, Koichiro Tanaka (Kyushu Sangyo University, Japan)
Pagepp. 307 - 310
KeywordFPGA, computer architecture, mobile application, education
AbstractIn our university, we take advantage of the educational microprocessor “KERNEL“ board for learning computer architecture. It is simply computer that operate by writing program. It is designed to be able to confirm internal state of CPU and memory. In our research, we develop new execution environment of KERNEL considering extensibility and portability. We implement the architecture of KERNEL on a FPGA board and an application of its console on a mobile device. We implement the function of transmitting internal running state of CPU to application with Bluetooth communication on the FPGA board and implement the writing instructions to memory and setting register value of KERNEL on the mobile application. In addition, we also implement the emulator function of KERNEL which run only application without FPGA board. This system can extend the function by rewriting FPGA and the application. In addition, students are able to learn computer architecture in their home.


Session M3-3  RF Circuits / Semiconductor Devices and Technology
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Room 3
Chair: Fujihiko Matsumoto (National Defense Academy of Japan, Japan)

M3-3-1 (Time: 15:50 - 16:10)
TitleA Cost-Efficient Tuner Design for Digital TV Receiver
AuthorShih-Chang Hsia, *Po-Yu Hsiao, Ming-Hwa Sheu, Hsin-Hsien Huang (National Yunlin University of Science and Technology, Taiwan)
Pagepp. 311 - 314
Keywordpolyphase, tuner, mixer, active filter
AbstractIn this study, a single-chip is designed with MOS techniques for the digital TV tuner. The chip consists of the modules of low noise amplifier (LNA), RF mixer, digital voltage control oscillator, and polyphase filter. The tuner is designed with the structure of a single conversion by mixing the Quadrature local signal with polyphase filter. The frequency band is designed from 530 MHz to 602 MHz for DTV channels in Taiwan. The gain of the entire tuner can be over 32 dBm, and the image rejection ratio (IRR) of the tuner is above 30 dBm. The local oscillator built-in digital-to-analog converter can be directly controlled by a digital code to select the TV channel. The silicon chip had been designed with a full-custom layout, where the chip size and core size is about 0.497 and 0.1095 mm2, respectively, when implemented by TSMC 0.18µm CMOS process. The maximum power dissipation is about 136.7 mW when the chip works 3.3V.

M3-3-2 (Time: 16:10 - 16:30)
TitleA Low Power Coupling Shielded Digitally Controlled Oscillator (DCO)
Author*Sang-Sun Yoo (Korea Advanced Institute of Science and Technology, Republic of Korea), Kang-Yoon Lee (Sungkyunkwan University, Republic of Korea), Hyung-Joun Yoo (Korea Advanced Institute of Science and Technology, Republic of Korea)
Pagepp. 315 - 317
KeywordDCO, coupling, shield ring, ADPLL, crosstalk
AbstractThis paper presents is a coupling shieled digitally controlled oscillator (DCO). The DCO has magnetic coupling shielding ring (CSR) which is designed with aluminum (Al) pad metal. This closed ring induces electromagnetic-force by Lenz’s Low. As a results, designed DCO can be protected from the attack of power amplifier (PA) with strong magnetic field. The DCO was fabricated in 55-nm CMOS process. By using CSR, isolation between DCO and PA is increased about 11 dB witout degradations of performances.

M3-3-3 (Time: 16:30 - 16:50)
TitleA Subthreshold Ultra-Low Power Low-Voltage High Resolution DCO using Al Pad metal Layers for BLE Application in 55 nm Technology
Author*Hamed Abbasizadeh, Seong-Jin Oh, Sang-Sun Yoo, Kang-Yoon Lee (Sungkyunkwan University, Republic of Korea)
Pagepp. 319 - 321
KeywordDCO, ultra-low power, high resolution
AbstractThis paper presents the design and implementation of an ultra-low power Digitally Controlled Oscillator (DCO) for Bluetooth Low-Energy (BLE) applications which has wide tuning range about 4.2 GHz ~ 6.1 GHz. The device consists of MOSFET circuits operating in the sub-threshold region and the power dissipation is reduced by sub-threshold design. High frequency resolution is achieved by using the fine tuning bank and the Delta-Sigma Modulator (DSM). A very small capacitance, about 16.22 aF through customized lateral MOM capacitor is utilized for attaining small switchable capacitance in fine tuning bank. To eliminate routing metals' parasitic inductance, a ultra-wide Aluminium (Al) pad metal layers are incorporated as a supper conductor. The proposed DCO is designed using CMOS 55 nm process. The S-parameter based simulation results prove that the current consumption is 480 µA at center frequency of 4.88 GHz at 0.55 V supply voltage. Also, high frequency resolution of 4.8 kHz is achieved. The DCO phase noise is about -114.1 dBc/Hz at 1-MHz offset. The layout size is 250×470 µm2

M3-3-4 (Time: 16:50 - 17:10)
TitleSelf-rectifying Resistive Switching Characteristics in Cu/IGZO/Si Structure
Author*Su-Hyun Bang, Sungjun Kim, Hyungjin Kim, Byung-Gook Park (Seoul National University, Republic of Korea)
Pagepp. 323 - 326
Keywordresistive switching, RRAM, self-rectyfing, IGZO
AbstractWe investigate the resistive switching and self-rectifying characteristics of proposed RRAM. We fabricate RRAMs using Cu/IGZO/Si structure. The bilayer film that consists of oxygen-deficient and oxygen-rich α-IGZO, and the monolayer film of oxygen deficient α-IGZO are fabricated by controlling oxygen concentration. Proposed RRAM cells show the typical hysteresis I–V curve including set and reset operation under the DC sweep mode. Furthermore, self-rectifying phenomenon is observed.

M3-3-5 (Time: 17:10 - 17:30)
TitleReconfigurable U-shape Tunnel FET
Author*Myounggon Kang (Korea National University of Transportation, Republic of Korea), Won Joo Lee, Yoon Kim (Pusan National University, Republic of Korea)
Pagepp. 327 - 330
Keywordreconfigurable FET, TFET
AbstractIn this paper, we propose and validate a novel design of dynamically reconfigurable tunnel field-effect transistor having a U-shape channel and three gates (two polarity gates and one control gate). The polarity gate acts on the side regions of the channel, swithcing the device polarity dynamically between n and p-type. Also, the conduction mechanism of proposed device is based on the tunnel FET operation that features low-leakage current, low power consumption, and scalable subthreshold swing (SS). To improve on-current drivablity and suppress short-channel effect, we propose a unique U-shape channel structure. The designed device provides ~30úI higher on-current(Ion) and an average subthreshold swing of 41.8 mV/dec.

M3-3-6 (Time: 17:30 - 17:50)
TitleAutonomous Neuromorphic System with Four-Terminal Si-Based Synaptic Devices
Author*Jungjin Park, Hyungjin Kim, Min-Woo Kwon, Sungmin Hwang, Byung-Gook Park (Seoul National University, Republic of Korea)
Pagepp. 331 - 334
Keywordneuromorphic, neuron circuit, synaptic device, action-potential, spike-timing-dependent-plasticity
AbstractWe have developed an autonomous neuromorphic system that can work with four-terminal Si-based synaptic devices. The symmetrical current mirrors connected to the n-channel synaptic devices constitute the synaptic connection and integration parts to express the excitation and the inhibition mechanism of neurons. The number and the weight of the synaptic devices affect the amount of the current flowing into the capacitor. The double-stage inverters controlling delay time and the NMOS with large threshold voltage (V_T) constitute the action-potential generation part. The generated action-potential is transmitted to the next neuron and simultaneously returned to the back gate of the synaptic device for expression of spike-timing-dependent plasticity.


Session M3-4  [SS4] Technologies for Beyond 4G Wireless Communications/Networks
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Room 4
Chair: Seokhyun Yoon (Dankook University, Republic of Korea)

M3-4-1 (Time: 15:50 - 16:10)
Title4/6 Modulation Code for Multi-level Holographic Data Storage
Author*Seongkwon Jeong, Oh-Soon Shin, Jaejin Lee (Soongsil University, Republic of Korea)
Pagepp. 335 - 338
Keywordholographic data storage, 2D-ISI, multi-level, modulation code
AbstractHolographic data storage system has good characteristics of high recording density, transfer rate, and short access time. On the other hand, it has a serious problem of two dimensional inter-symbol interference. Furthermore, because one pixel can save more than 1 bit per pixel in multi-level HDS, the storage system has small noise margin for symbol detection that causes more errors. In this paper, we propose a 4/6 modulation code which mitigates two dimensional inter-symbol interference for multi-level HDS.

M3-4-2 (Time: 16:10 - 16:30)
TitleUsing Density Evolution for Analysis of Low Complexity MIMO Detector based on Belief Propagation over a Ring Type Pair-wise Graph
Author*Seokhyun Yoon (Dankook University, Republic of Korea)
Pagepp. 339 - 340
KeywordMIMO, Belief propagation, pair-wise graph
AbstractThe convergence and density evolution of a low complexity MIMO detection based on belief propagation over a ring-type pair-wise graph is considered for binary input. The algorithm has been originally proposed in [1], where the convergence for Gaussian input have been analyzed. Here, we extend the convergence analysis to binary data and provide an asymptotic performance in terms of BER and SINR via density evolution analysis.

M3-4-3 (Time: 16:30 - 16:50)
TitleAnalysis of Throughput Achieved by Contention-based MAC Scheme in Wireless Passive Sensor Networks: Case Study
AuthorHeewon Seo, Jun Ha, Jin Kyung Park, *Cheon Won Choi (Dankook University, Republic of Korea)
Pagepp. 341 - 344
Keywordwireless passive sensor network, MAC scheme, framed and slotted ALOHA, throughput formula, capture
AbstractDifferent from a conventional wireless sensor network, a wireless passive sensor network distinctively has RF sources, which feed energy to sensor nodes by radiating RF waves. Against theoretical expectations about an eternal life, a wireless passive sensor network suffers from many practical difficulties; scarcity of energy, non-simultaneity of energy reception and data transmission and inefficiency in allocating time resource. Perceiving such difficulties, we consider a contention-based MAC scheme, which is rooted in framed and slotted ALOHA, for sensor nodes to transmit packets to a sink node. Then, we analytically investigate the network-wide throughput attained by the MAC scheme when the sensor nodes are randomly scattered in the initial phase and the transmitted packets experience path losses. Especially, in a network consisting of 3 sensor nodes, we derive the exact throughput formula in a closed form. The numerical examples produced by the throughput formula confirm that the contention-based MAC scheme can be optimized as to achieve maximum throughput by properly deciding the length of an acting interval.

M3-4-4 (Time: 16:50 - 17:10)
TitleAntenna Selection for Full-Duplex MIMO Systems
AuthorSeokju Jang, Hoon Lee, *Inkyu Lee (Korea University, Republic of Korea)
Pagepp. 345 - 347
KeywordMIMO, Full-duplex, Antenna selection
AbstractIn this paper, we investigate an antenna selection (AS) method for full-duplex multiple-input multiple-output systems, in which antennas at each node can be selected to either transmit or receive. In this configuration, we analyze the average sum rate of the optimal AS scheme which finds the best antenna set solution by examining all the possible candidates. The result provides insight on which transmit and receive antenna configuration improves the average sum rate performance of the AS scheme. From simulation results, we verify that our analysis matches the numerical results, and a sub-optimal AS scheme which chooses an atenna set maximizing the sum rate under the derived antenna configuration provides a near-optimal performance.

M3-4-5 (Time: 17:10 - 17:30)
TitleInter-Cell Interference Coordination for Multimedia Communications
Author*Seok-Ho Chang (Dankook University, Republic of Korea)
Pagepp. 349 - 350
KeywordInter-Cell Interference, Coordination, Multimedia Communications
AbstractIn this paper, we study the inter-cell interference coordination for the transmission of progressive images which need unequal spectral efficiencies in the bitstream. Based on our analysis for the outage probabilities of the full and partial frequency reuses, we suggest the use of fractional frequency reuse for the transmission of a series of progressive packets.

M3-4-6 (Time: 17:30 - 17:50)
TitleA Self-Organizing Wireless Sensor Network for Industrial Monitoring
Author*Zafar Iqbal, Heung-No Lee (Gwangju Institute of Science and Technology, Republic of Korea)
Pagepp. 351 - 354
Keywordindustrial wireless sensor networks, cooperative communication, environment monitoring, fusion, self-organizing network
AbstractRecently, industrial wireless sensor networks (IWSNs) have been widely used for monitoring the state of heavy machinery and large factories whose condition is critical to the operation of machine as well as the safety of people around them. However, the wireless communication channels, which these sensors must operate in, are not always reliable, and observations of some sensors cannot be reported successfully to the base station. In order to deal with this problem, we propose a cooperative WSN scheme, which organizes itself into groups of cooperation nodes and then start normal operation. The proposed scheme effectively increases the probability of correct decision about the state of the industrial area, and reduces the probability of false alarms at a given signal level. We also propose a novel cooperation protocol that uses the minimum number of transmissions for delivering the cooperative sensing information to the base station, thereby, reducing the overall traffic and energy usage of the network.


Session M3-5  Image Processing (3)
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Room 5
Chair: Jong-Il Park (Hanyang University, Republic of Korea)

M3-5-1 (Time: 15:50 - 16:10)
TitleAdaptive Mode Selection for Low Complexity Enhancement Layer Encoding of SHVC
Author*Kazuki Kuroda, Takafumi Katayama, Tian Song, Takashi Shimamoto (The University of Tokushima, Japan)
Pagepp. 355 - 358
KeywordSHVC, Merge mode, Motion estimation, Complexity reduction, Coding efficiency
AbstractIn this work, we proposed a low complexity algorithm using the adaptive mode selection based on the sum of squared errors of the correlated pixels and motion vector. To achieve low computation complexity, the temporal correlation and the spatial correlation is used for the mode decision. The proposed algorithm is evaluated by the reference software of SHVC. The simulation results show that the proposed algorithm can achieve over 20% computation complexity reduction comparing to the original SHVC algorithm.

M3-5-2 (Time: 16:10 - 16:30)
TitleProposal of Haze Removal Method with Adjustable Processing Degree
Author*Yi Ru, Go Tanaka (Nagoya City University, Japan)
Pagepp. 359 - 362
Keywordhaze removal, atmospheric scattering, bilateral filter
AbstractImages are often degraded by haze, and the visibility of an image which includes haze is poor. To improve the visibility of such images, some methods have been proposed. In this paper, we propose a new haze removal method which can easily adjust the processing degree by two parameters. It enables avoiding the excessive and inadequate processing problems. In addition, the haze residue in partial area is suppressed in the proposed method. We also use the bilateral filter to sharpen an output image. Experimental results show the effectiveness of the proposed method.

M3-5-3 (Time: 16:30 - 16:50)
TitleExemplar-Based Image Inpainting Using Context-Aware Approach
Author*Yusuke Murakami, Koichi Ichige (Yokohama National University, Japan)
Pagepp. 363 - 366
KeywordExemplar-based image inpainting, Context-aware approach
AbstractThis paper presents a novel exemplar-based inpainting method using context-aware approach. There exist conventional methods which use the positional relations between the most similar patches in patch unit, but it cannot achieve the best performance. By applying context-aware approach to it, we propose an high-quality and efficient exemplar-based inpainting method. Performance of the proposed method is evaluated through computer simulation.

M3-5-4 (Time: 16:50 - 17:10)
TitleAccurate Feature Point Tracking for Omnidirectional Image Sequence
Author*Hiroshi Tada, Koichi Ichige (Yokohama National University, Japan)
Pagepp. 367 - 370
Keywordfeature point tracking, omnidirectional image, optical flow
AbstractIn this paper, we propose a highly accurate feature point tracking method for omnidirectional image sequence. First we try to enlarge the number of optical flows by interpolation. Then we improve the tracking accuracy by combining the flows which will make smaller tracking error. Performance of the proposed method is evaluated through computer simulation.

M3-5-5 (Time: 17:10 - 17:30)
TitleLocal Signal-Dependent Noise Estimation on Texture Domain for CFA Raw Images
Author*Kyu-Ho Lee, Jong-Ok Kim (Korea University, Republic of Korea)
Pagepp. 371 - 374
Keywordimage denoising, signal-dependent noise, local noise estimation, CFA raw image
AbstractIn this paper, we propose a method of signal-dependent noise estimation and denoising that operates on the CFA raw image. The proposed method effectively deals with signal dependent noise by estimating and denoising noise on the texture domain in a CFA LR component-wise and localized manner. The proposed method is practically evaluated on a simulated end-to-end imaging pipeline. The experimental results indicate that the proposed method indeed efficiently removes signal-dependent noise.

M3-5-6 (Time: 17:30 - 17:50)
TitleEmbedded Implementation Model of 2-D Non-separable Oversampled Lapped Transforms for Video Processing
Author*Keita Imai, Shogo Muramatsu (Niigata University, Japan)
Pagepp. 375 - 378
KeywordEmbedded implementation, Non-separable oversampled lapped transform, HW/SW co-implementation, Image processing
AbstractThis work proposes an embedded implementation model of 2-D Non-separable Oversampled Lapped Transform (NSOLT) for video processing. NSOLT has successfully been applied to image restoration. The authors have proposed an embedded implementation model of NSOLT as a previous work. The existing model,however, cannot efficiently process successive frames of a video. To realize efficient video processing, a modified embedded implementation model of NSOLT is proposed by introducing a pipeline architecture.


Session M3-6  Computer Systems & Application (3)
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Room 6
Chair: Kyung Youn Kim (Jeju National University, Republic of Korea)

M3-6-1 (Time: 15:50 - 16:10)
TitleA Control Interface of Level of Immersion in VR System
Author*Rina Ashida, Mitsunori Makino (Chuo University, Japan)
Pagepp. 379 - 382
KeywordVirtual Reality, Head Mount Display
AbstractIn this paper, an adjuster for virtual reality (VR) environment is proposed, with which a user can control level of visual immersion by him/herself. The dial-type input interface is suitable for the user who wears head-mounted-display (HMD) to manipulate easily without interference on his/her view.

M3-6-2 (Time: 16:10 - 16:30)
TitleAn Interactively Virtual Extended Display on Transparent HMD
Author*Daiki Masaki, Mitsunori Makino (Chuo University, Japan)
Pagepp. 383 - 386
KeywordAR, HMD
AbstractIn this paper, an AR-based expansion of PC display is proposed, which makes us work more effectively with multi windows. Wearing a transparent head-mounted-display, a user can see three virtual extended screens of PC, as well as an actual screen. When he/she bobs or weaves his/her head, the virtual screens are correspondingly displayed/hided on the HMD. The feature contributes to preventing the virtual screens from obstructing his/her view. The result of user-test shows usefulness of the proposed system.

M3-6-3 (Time: 16:30 - 16:50)
TitleCompact and High-Speed Hardware Feature Extraction Accelerator for Dense Scale-Invariant Feature Transform
Author*Tetsushi Koide, Takumi Okamoto, Tatsuya Shimizu, Koki Sugi, Anh-Tuan Hoang, Toru Tamaki, Bisser Raytchev, Kazufumi Kaneda (Hiroshima University, Japan), Shigeto Yoshida, Hiroshi Mieno (Hiroshima General Hospital of West Japan Railway Company, Japan), Shinji Tanaka (Department of Endoscopy and Medicine Graduate School of Biomedical and Health Science, Hiroshima University, Japan)
Pagepp. 387 - 390
KeywordComputer-Aided Diagnosis, Endoscopic Image Processing, FPGA, Dense Scale-Invariant Feature Transform (D-SIFT), Real-Time Processing
AbstractThis paper presents a D-SIFT based feature extraction hardware accelarator in real-time computer-aided diagnosis (CAD) system for endoscopic images. From an FPGA implimentation, we have demonstrated the proposed hardware oriented D-SIFT architecture was very compact due to no multipication and was very suitable for stream based image processing. The processing speed for Full-HD (1920x1080) high resoltion image is only 20 msec@100 MHz and it is about 700 times faster than that of software implimentation (14 sec) within low latency. The proposed D-SIFT accelarator can be also applicable for the feature exstruction part for various types of image processing including 4 K and 8 K high resolusion images.

M3-6-4 (Time: 16:50 - 17:10)
TitleParallel Particle Swarm Optimization on Many-Integrated-Cores Platforms
AuthorYasuyuki Miyazato (University of the Ryukyus, Japan), *Takeshi Tengan (Meio University, Japan), Andrea Veronica Porco, Morikazu Nakamura (University of the Ryukyus, Japan)
Pagepp. 391 - 394
KeywordParallel Processing, PSO, Optimization
AbstractThis paper presents parallel particle swarm optimization (PSO) on Many-Integrated-Cores (MIC) architecture platforms and investigate how to utilize many cores of MIC platforms efficiently. Experimental results show the relation between searching performance and search strategies in the parallel PSO. We confirm that the balance between the width and depth in search strategies is essential for efficient searching.

M3-6-5 (Time: 17:10 - 17:30)
TitleTreatments of Children with Autism Spectrum Disorders using a Spherical Robot with Facial Expressions
Author*San Ratanasanya, Anon Leekphai, Thanat Kitisook, Boonserm Kaewkamnerdpong, Jumpol Polvichai (King Mongkut's University of Technology Thonburi, Thailand)
Pagepp. 395 - 398
KeywordAutism, Autism Spectrum Disorders, Embedded System, Robot with Facial Expressions, Spherical Robot
AbstractAutism Spectrum Disorder (ASD) is a condition in many children, which referred to a broad set of behavioral disorders. Typically, children with ASD cannot communicate and interact with people, and as a result, having no social activities at all. Fortunately, they can interact with non-human object such as robots. There are hence many researches to treat these children using different shapes and types of robot. This paper investigates the effects of the spherical robot with facial epressions and movements in treating children with ASD. The designs, both structural and system, and suitable features of robot are studied and proposed. The preliminary experiments show impressive results that the proposed designs and features of the robot can successfully draw attention from 75% of the patients with different levels of disorders. However, more cases should be experimented in order to get more suitable details for treating ASD patients.



Tuesday, July 12, 2016

Session T1-1  Mobile & Wireless Communications (4)
Time: 9:00 - 10:20 Tuesday, July 12, 2016
Location: Room 1
Chair: Poompat Saengudomlert (Bangkok University, Thailand)

T1-1-1 (Time: 9:00 - 9:20)
TitleClosed form of Spectral Efficiency for Non Orthogonal Multiple Access in Nakagami Fading Environment
AuthorPongsatorn Sedtheetorn (Mahidol University, Thailand), *Kritsada Mamat (Kasetsart University, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 399 - 402
KeywordNOMA
AbstractIn this paper, an original analysis on the uplink spectral efficiency of non-orthogonal multiple access (NOMA) is introduced. Based on our accurate approximation, a closed-form expression of the spectral efficiency is proposed. Unlike the literature, the proposed expression yields accurate values, validated by Monte- Carlo simulation, of the spectral efficiency when the channel gains are Nakagami distributed and the random number of active users are taken into account.

T1-1-2 (Time: 9:20 - 9:40)
TitleBlock-based Transmission with Adaptive Reliability Control for Ad Hoc Networks
Author*Tomoaki Sakaguchi, Taku Yamazaki (Waseda University, Japan), Ryo Yamamoto (The University of Electro-Communications, Japan), Yoshiaki Tanaka (Waseda University, Japan)
Pagepp. 403 - 406
KeywordWireless, Ad hoc network, Transport protocol, Reliability
AbstractIn ad hoc networks, each node communicates with other nodes directly or with multi-hop transmission. However, a packet loss rate between a source node and a destination node tends to be high due to mobility of nodes or radio interferences. Therefore, conventional transport protocols that are used in wired network cannot perform their potentials in ad hoc network since they employ an end-to-end acknowledgement. In this paper, we propose an adaptive reliability control mechanism for ad hoc network and clarify the performance with the computer simulation.

T1-1-3 (Time: 9:40 - 10:00)
TitleInter-Cell Interference Suppression at Wireless Vehicle Access Networks with Distributed Subcarrier Mapped OFDM
Author*Garam Yu, Haesoon Lee, Joonki Kim, Daesik Hong (Yonsei University, Republic of Korea)
Pagepp. 407 - 409
Keywordvehicle access network, vehicular network, V2X, OFDM, inter cell interference
AbstractWe propose a ICI suppression technique by using the Distributed Subcarrier Mapped Orthogonal Freqeuncy Division Multiplexing (DS-OFDM) for wireless vehicle access networks. Compared to the Localized Subcarrier Mapped OFDM (LS-OFDM), DS-OFDM has a merit of reducing the intensity of inter-cell interference in fast moving vehicle circumstances thanks to Doppler spreading effect. Numerical results show that DS-OFDM outperforms LS-OFDM in terms of outage probability espeically when we set the threshold SINR lower than the average of SINR.

T1-1-4 (Time: 10:00 - 10:20)
TitleFPGA Implementation of Multiprocessor SoC for Industrial Wireless LAN Systems
Author*Yushi Matsunaga, Mitsuru Hamada, Nico Surantha, Yuhei Nagao, Masayuki Kurosaki, Hiroshi Ochi (Kyushu Institute of Technology, Japan)
Pagepp. 411 - 414
Keywordindustrial Wireless Local Area Network, multiprocessor, SoC, AMP
AbstractIn this work, an industrial Wireless Local Area Network system used to control industrial robots in factory automation (FA) environments is addressed.For real-time property, we propose a multiprocessor architecture. A multiprocessor architecture can support real-time property. Therefore, our proposed schemes can achieve fast and safe performance for FA communication systems.


Session T1-2  Analog Circuits (1)
Time: 9:00 - 10:20 Tuesday, July 12, 2016
Location: Room 2
Chair: Katsutoshi Saeki (Nihon University, Japan)

T1-2-1 (Time: 9:00 - 9:20)
TitleElectronically Tunable Multiple-Input Single-Output Voltage-Mode Multifunction Filter Using OTA and DURC
Author*Ukrit Kornkanok, Paitoon Rakluea, Virote Pirajnanchai (Rajamangala University of Technology Thanyaburi, Thailand), Supachai Klungtong (Kasetsart University, Thailand)
Pagepp. 415 - 418
KeywordDURC, OTA, voltage-mode, multifunction filters, low power
AbstractThis paper presents the design method for multifunction filter circuit. The characteristics of filter (low-pass, high-pass, band-pass, band reject filter, and all-pass) can be obtained without changing circuit topology. Its scheme is principally composed of passive element using double capacitive layers uniform distributed RC (DURC) and active element using operational transconductance amplifier (OTA). The frequency response of filter circuit can be adjusted by the transconductance gains of OTA without affecting the circuit’s quality factor (Q). The performance of the proposed filter circuit was simulated by PSpice. Its simulation results were performed for the low-voltage power supply with only plus/minus 2VDC. In addition, the circuit is advantage that there are good magnitude response and low sensitivities of less than one unit. The simulation results of circuit are in good agreement with the theoretical calculations. The proposed filter circuit is very suitable to further develop into a VLSI circuit for communication and signal processing system.

T1-2-2 (Time: 9:20 - 9:40)
TitleSynthesis of a First-order Passive Complex Coefficient Bandpass Filter Including No Transformers
Author*Tatsuya Fujii, Kazuhiro Shouno (University of Tsukuba, Japan)
Pagepp. 419 - 422
Keywordcomplex, analog, filter, inductor, bandpass
AbstractIn this paper, a method for designing a first-order passive complex bandpass filter including no transformers is proposed. The proposed frequency transformation is obtained by arranging the conventional one for the complex filter using loose coupling transformers. As a result, we can exclude transformers by giving appropriate specifications to the conventional frequency transformation. As an example, a complex bandpass filter obtained from a prototype lowpass filter is designed. The validity of the proposed method is confirmed through computer simulation.

T1-2-3 (Time: 9:40 - 10:00)
TitleA Ratio-insensitive Switched-Capacitor Algorithmic Digital-to-Analog Converter Using Sample/Hold and Divider
Author*Hiroki Matsumoto (University of Miyazaki, Japan)
Pagepp. 423 - 426
KeywordSwitched-capacitor, DA converter, Capacitance mismatch, Sample/hold, Divider
AbstractA novel switched-capacitor algorithmic digital-to-analog (D/A)converter using sample/hold and divider is described. The operation is insensitive to capacitor mismatch. Thus, accurate D/A conversion is possible on standard CMOS process technology. Accuracy estimation shows that 15 bit resolution is possible on capacitor mismatch of 1 percent.


Session T1-3  Graphs and Net Theory
Time: 9:00 - 10:20 Tuesday, July 12, 2016
Location: Room 3
Chair: Yousuke Tanigawa (Osaka Prefecture University, Japan)

T1-3-1 (Time: 9:00 - 9:20)
TitleA New Approach to Solve the Partitioning Problem to Realize Highly Reliable Telecommunications Networks
AuthorTakayuki Saito, *Masahiro Hayashi (Tokyo City University, Japan)
Pagepp. 427 - 430
KeywordTelecommunications Networks, Reliability, Partitioning Problem, Delta-star Transformation, Factoring Algorithm
AbstractWe propose a new algorithm to solve the partitioning problem. The problem is to check whether a telecommunication network can be divided into two subnetworks such that we can ensure the independence of two paths by assigning these paths in different subnetworks. The basic idea of our proposed algorithm is to transform a special type of configuration called delta to simplify the network topology. Numerical examples show that it realizes actually faster execution time for checking rather than previous algorithm.

T1-3-2 (Time: 9:20 - 9:40)
TitleAnalysis of Cascading Failure Rate for Telecommunications Networks
AuthorTakuya Omura, *Masahiro Hayashi (Tokyo City University, Japan)
Pagepp. 431 - 434
KeywordCascading failures, Telecommunications Networks, Reliability, Survival Traffic Rate
AbstractThis paper proposes a new measure of cascading failures in telecommunications networks, where cascading failures are wide area damages caused by a small size triger failure. Previous works do not consider the probability of such cascading failures appear. Now this paper propose a new measure to evaluate such probability. Key idea is that we do not use traditiona reliablity measure availablity because it needs information of repair time while such repair tiime is difficult to estimate in real situation. Our measure cascadifng failure rate is defined as the inverse of mean time to first cascading failure, where we do not need information of repair time.

T1-3-3 (Time: 9:40 - 10:00)
TitleThe Shortest Routes of Two UAVs on Square Grid Graph with Boundary Depot
Author*Yoshihiro Kaneko, Takeki Kani (Gifu University, Japan), Kenichi Mase (Niigata University, Japan)
Pagepp. 435 - 438
Keywordunmanned aerial vehicle, the shortest routes, Hamilton cycle, grid graph, boundary depot
AbstractAn unmanned aerial vehicle UAV surveillance system is composed of some pairs of electric helicopter EH and electric vehicle EV to survey a designated area. Each EH and EV pair surveys one assigned sub-area within limited time. So far we have shown the shortest twice flight routes of EH in sub-area with corner depot. The paper relaxes such depot condition and deal with the boundary depot location. It turns out for such EHs, the shortest flight routes has length at most(n+7)/2, where n is the number of surveillance units of sub-area.

T1-3-4 (Time: 10:00 - 10:20)
TitleModeling and Analysing Boolean Networks by Coloured Petri Nets
Author*Giang Van Trinh (Department of Software Engineering, Ho Chi Minh City University of Technology, Viet Nam), Hiraishi Kunihiko (School of Information Science, Japan Advanced Institute of Science and Technology, Japan), Tho Thanh Quan (Department of Software Engineering, Ho Chi Minh City University of Technology, Viet Nam)
Pagepp. 439 - 442
KeywordBoolean networks, Attractors, Petri nets, Coloured Petri nets
AbstractWe propose a new method for modeling and analysing Boolean networks based on Coloured Petri nets. All synchronous, asynchronous, and probabilistic dynamics are expressed in the proposed method. We demonstrate the advantage of our approach in the issue of attractor detection, which is an important long-term behaviour of Boolean networks.


Session T1-4  Computers
Time: 9:00 - 10:20 Tuesday, July 12, 2016
Location: Room 4
Chair: Ittetsu Taniguchi (Ritsumeikan University, Japan)

T1-4-1 (Time: 9:00 - 9:20)
TitleAn OpenCL Framework for FPGA-based Heterogeneous Multicore Architecture
Author*Shunsuke Takai, Ittetsu Taniguchi, Hiroyuki Tomiyama (Ritsumeikan University, Japan), Sri Parameswaran (University of New South Wales, Australia)
Pagepp. 443 - 444
Keywordheterogeneous multicores multicore, OpenCL, FPGA
AbstractThis paper presents an OpenCL framework for heterogeneous multicore architecture on an FPGA. The architecture consists of two Cortex-A9 hard-cores and four Nios-2 soft-cores, and our OpenCL framework performs load balancing on the two different kinds of cores. In the preliminary experiments, we have tested our OpenCL framework on their cores and confirmed the our framework works correctly.

T1-4-2 (Time: 9:20 - 9:40)
TitleILP-based Scheduling for Malleable Parallel Tasks
Author*Kana Shimada (Ritsumeikan University, Japan), Shogo Kitano (The University of Electro-Communications, Japan), Ittetsu Taniguchi, Hiroyuki Tomiyama (Ritsumeikan University, Japan)
Pagepp. 445 - 446
Keywordtask scheduling, integer linear programming
AbstractTask scheduling is one of the most important processes in the design of multicore computing systems. This paper presents a technique for scheduling of malleable tasks. Our scheduling technique decides not only the execution order of the tasks but also the number of cores assigned to the individual tasks, simultaneously. We formulate the scheduling problem as an integer linear programming (ILP) problem, and the optimal schedule can be obtained by solving the ILP problem. Experiments using a standard task-set suite clarify both the strength and weakness of this work.

T1-4-3 (Time: 9:40 - 10:00)
TitleProposal for the Trajectory Control of Two Wheels Mobile Robot with Object Pushing and Self-position Estimation
Author*Ayato Ueda, Yoshihiro Yasutake, Sunao Sawada (Kyushu Sangyo University, Japan)
Pagepp. 447 - 450
KeywordMobile Robot, Position Control, Trajectory Control, Spline Curve, Push-a-box Operation
AbstractAutonomous mobile robot has two wheels and travels while pushing object in the fixed arm. It is passing through a specified point by using the self-position estimation. We propose a method of orbit control to put object to the point of the target.


Session T1-5  Computer Vision (1)
Time: 9:00 - 10:20 Tuesday, July 12, 2016
Location: Room 5
Chair: Jonk-Ok Kim (Korea University, Republic of Korea)

T1-5-1 (Time: 9:00 - 9:20)
TitleFast UHD Background Modelling with Mixed Order Block Updates
Author*Axel Beaugendre, Satoshi Goto, Takeshi Yoshimura (IPSRC, Waseda University, Japan)
Pagepp. 451 - 454
KeywordUHD, 4K, Background Modelling, Detection, MBBM
AbstractUp to now, the majority of the background modelling methods are pixel-based and rely heavily on hardware optimizations to process standard definition videos in real-time. There are far too complex and require too much resources (time, memory, etc.) to be applied to Ultra High Definition (UHD) videos. We propose a fast background modelling method which divides the complexity through time via the update of multiple small blocks. By using mix of linear and random selecting order we insure that all blocks get updated in a known number of frames. Preliminary results show that we can reduce by at 80 times the modelling time compared to a full frame modelling and reach an average processing speed of 77 fps for 4K videos.

T1-5-2 (Time: 9:20 - 9:40)
TitleNear Real-Time Tracking System for 8K UHD Videos
Author*Axel Beaugendre, Satoshi Goto, Takeshi Yoshimura (IPSRC, Waseda University, Japan)
Pagepp. 455 - 458
KeywordUHD, 8K, Tracking, Real-Time
AbstractPixel based background subtraction methods and appearance-based tracking systems are too complex and too slow to be applied to UHD. We present here a near real-time tracking system for 8K UHD videos which takes advantage of a block-propagative background subtraction method, an object refinement method and an object-based Hungarian tracking method. By combining the three methods, it is possible to track objects in an 8K UHD video with an average speed of about 6.5 fps including detection and tracking time.

T1-5-3 (Time: 9:40 - 10:00)
TitleScene Text Localization Using A Two-Class Detector Constructed By Filtered Channel Features
Author*Takuro Oki, Ryusuke Miyamoto (Meiji University, Japan)
Pagepp. 459 - 462
Keywordscene text localization, filtered channel features, sliding window
AbstractThe authors have tried to improve both accuracy and processing speed for scene text localization using integral channel features for construction of detector in the previous work but sufficient accuracy did not achieved. To improve the detection accuracy of the framework, this paper tries to apply filtered channel features that shows excellent and much better performance than deep learning for human detection.

T1-5-4 (Time: 10:00 - 10:20)
TitleDefect Classification of Electronic Board Using Bag of Features and Color Information
Author*Hidenobu Inoue, Yuji Iwahori (Chubu University, Japan), Boonserm Kijsirikul (Chulalongkorn University, Thailand), Manas K. Bhuyan (Indian Institute of Technology Guwahati, India)
Pagepp. 463 - 466
KeywordDefect Classification, Bag of Features, Color, SVM
AbstractThis paper proposes a new approach of defect classification using Bag of Features and color information in order to correspond to the various defect images without using any reference images. The purpose of the paper is to classify the true defect and pseudo defect such as dust on the electronic board. First, features are extracted from each image of data set and histogram features are generated and represented by Bag of Features. Next, noise removal is applied and combined features which consist of Bag of Features and color information are used. After extracting features, SVM (Support Vector Machine) is used for the learning and classification. The usefulness of the proposed approach is confirmed by evaluating the accuracy of defect classification in comparison with the previous approaches with the target images of electronic board images which includes the actual defects.


Session T2-1  Mobile & Wireless Communications (5)
Time: 13:30 - 15:30 Tuesday, July 12, 2016
Location: Room 1
Chairs: Kritsada Mamut (Kasetsart University, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)

T2-1-1 (Time: 13:30 - 13:50)
TitleOptimal Estimation Interval for Time-Varying V-BLAST Channels
AuthorPisit Vanichchanunt (King Mongkut's University of Technology North Bangkok, Thailand), *Kritsada Mamat (Kasetsart University, Thailand), Panupat Wipaweeponkul, Thousapol Thitivorakarn, Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 467 - 470
KeywordV-BLAST
AbstractIn this paper, we study an effect of channel estimation error on time-varying V-BLAST system. A first-order auto regressive (AR) is introduced to model the time evolution characteristic. The receiver can obtain the current channel information by using a pilot symbol. The estimation accuracy depends on amount of pilot symbols. Based on the received signal, the receiver uses an estimated channel matrix to detect the transmit symbol where Maximum Likelihood (ML), Zero-Forcing (ZF) and Minimum Mean Square Error (MMSE) are applied. For given total amount of pilot symbol and interested time period, we propose to divide all channels into equal intervals or groups, which each group uses the same estimated channel. We also specify an optimal number of groups and show that the optimal number of groups mainly depends on time correlation parameter and training budget. Numerical results show that the system can achieve the performance closest to perfect channel estimation when operating on a suitable estimation interval.

T2-1-2 (Time: 13:50 - 14:10)
TitleStudy of Idle skipped Dynamic Frame Slotted ALOHA for RFID Systems
Author*Sanika Krishnamali Wijayasekara, Chairat Phongphanphanee (Department of Electrical Engineering Chulalongkorn University, Thailand), Robithoh Annur (Department of Computer and Communications Technology, Tunku Abdul Rahman University, Malaysia), Suvit Nakpeerayuth (Chulalongkorn University, Thailand), Lunchakorn Wuttisittikulkij (Department of Electrical Engineering Chulalongkorn University, Thailand)
Pagepp. 471 - 474
KeywordDFSA
AbstractThe conventional dynamic frame slotted aloha algorithm (DFSA) adjusts the frame size according to the number of tags and continuously process until all the tags are identified. In this study we analyze the effect of skipping slots which are not carrying any information in order to enhance the performances of traditional DFSA. In addition, to compare the performances of idle skipped DFSA algorithm with conventional DFSA algorithm based on the mean number of slots require to identify all the tags. The mathematical analysis and simulation results show that the presented idle skipped DFSA algorithm is more effective when compared to the conventional DFSA algorithm.

T2-1-3 (Time: 14:10 - 14:30)
TitleHybrid Model RFID Anti-collision Algorithm
Author*Robithoh Annur (Universiti Tunku Abdul Rahman, Malaysia), Suvit Nakpeerayuth, Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 475 - 478
KeywordRFID, Anti-collision, Aloha, Tree Algorithm
AbstractThis paper addresses the problem of tags collision in an RFID network. We proposed a simple mechanism that includes framed slotted Aloha and tree algorithm. The framed slotted Aloha is used to estimate the number of collided tags. The process continues with the dynamic tree algorithm by splitting the collided tags into the number estimated tags. The results show that the frame size has significant effect to the system efficiency. The maximum system efficiency achieved by the proposed algorithm is 40.03%.

T2-1-4 (Time: 14:30 - 14:50)
TitleA STA initiated MAC Protocol for IEEE 802.11 ax Uplink Multi-User Communication
Author*Takuma Miyamoto, Yuichi Sugiyama, Tatsumi Uwai, Yuhei Nagao, Leonardo Jr. Lanante, Hiroshi Ochi (Kyushu Institute of Technology, Japan)
Pagepp. 479 - 482
KeywordWireless LAN, MAC Protocol, 11ax, Uplink Multi-User
AbstractIn this paper, we proposed a station (STA) initiated protocol for random uplink multi-user access in accordance with the 802.11ax wireless standard. In contrast to access point (AP) initiated protocols which is difficult to implement because it requires knowledge of transmission demand and timing, STA initiated protocols does not need any prior information as long as STAs use RTS frames. As a result of the simulation, the proposed method better than the AP initiated protocol.


Session T2-2  Analog Circuits (2)
Time: 13:30 - 15:30 Tuesday, July 12, 2016
Location: Room 2
Chair: Robithoh Annur (Tunku Abdul Rahman University, Malaysia)

T2-2-1 (Time: 13:30 - 13:50)
TitleImprovement Techniques of Low Frequency Characteristic for a Symmetrical Floating Impedance Scaling Circuit
AuthorFujihiko Matsumoto, Syuzo Nishioka, *Shota Matsuo, Takeshi Ohbuchi (National Defense Academy of Japan, Japan)
Pagepp. 483 - 486
KeywordImpedance scaling, CMOS, Low frequency, Active filters
AbstractImpedance scaling is a technique to multiply apparent capacitance using current feedback. The authors have proposed a symmetrical floating impedance scaling (SFIS) circuit. The conventional SFIS circuit has cascode stages to enhance resistance components for low frequency characteristics. However, the improvement of low frequency characteristics by cascode is not sufficient for a few Hz applications. This paper presents two methods to improve the low frequency characteristics of the SFIC circuit.

T2-2-2 (Time: 13:50 - 14:10)
TitleSelf-clock generating TDC for Single-Slope Analog-to-Digital Converter
Author*Moung-Sub Bea, Jin-Tae Kim, Lee Jong Cheol, Jang-Kyoo Shin, Pyung Choi (Kyungpook National Unversity, Republic of Korea)
Pagepp. 487 - 490
KeywordTime to Digital Converter, Ring oscillator, Analog to digital Converter
AbstractWe propose that Time to Digital converter (TDC) and Ring oscillator are unified into one circuit. And encoder is modified for proper operation. The structure give space efficiency. By Unifing two circuit and modified encoder. The encoder also provide relability. Our work is based on 0.35um process and Virtuoso schematic simulator. Proposed circuit consist of 5-bit TDC and 8-bit counter.

T2-2-3 (Time: 14:10 - 14:30)
TitleA Design of High Sensitivity Bridge Circuit using only Two Differential Resistive Sensors
Author*Jin-su Kim, Yui-hwan Sa, Ronnie Serfa Juan, Hyeong-woo Cha, Hi-seok Kim (Cheongju University, Republic of Korea)
Pagepp. 491 - 494
Keywordanalog circuits, resistive sensor, voltage-to-current converter, CMOS Op-amp
AbstractA novel resistance difference to voltage conversion (ΔRVC) circuit that uses a linear CMOS voltage to current converter (VIC) and an operational amplifier (op-amp) with a low cost and high accuracy measurement system is designed. This proposed circuit was designed for one-chip fabrication and consists of a VIC, an op-amp and a two resistive sensors. The main characteristic of this circuit configuration does need any matching for resistive element and has an advantage of twice the sensitivity compare to a conventional bridge circuit. Simulation results show that the resistance difference to voltage conversion circuits agree with theoretical equation.

T2-2-4 (Time: 14:30 - 14:50)
TitleA 1V/1.8V Ultra Low Noise LDO for 5.6GHz PLL Applications
Author*Thi Kim Nga Truong, Hamed Abbasizadeh, Sung-Hun Cho, Dong-Soo Lee, Kang-Yoon Lee (Sungkyunkwan University, Republic of Korea)
Pagepp. 495 - 497
Keywordlow noise, LDO
AbstractThis paper presents a low noise low drop voltage (LDO) for 5.6GHz PLL application. The proposed LDO employs internal RC-Filter with multi-layer capacitor at the output of the bandgap reference (BGR) to achieve ultra-low noise at interest frequencies. It is also a mode changeable LDO so that the output voltage level can be 1V/1.8V. The 4-bits controlled resistor ladder is adopted to compensate the process voltage temperature (PVT) variation. The highest output noise of the LDO are 5.68fV2/Hz and 23fV2/Hz at 10 kHz in 1V/1.8V mode respectively. The proposed LDO is implemented in CMOS 65nm technology with the die size is 460um x 290um.

T2-2-5 (Time: 14:50 - 15:10)
TitleImplementation of Low Power Consumption of Neuromorphic Devices
Author*Takahiro Toizumi, Yoshiki Sasaki, Katsutoshi Saeki (Nihon University, Japan)
Pagepp. 499 - 502
KeywordANN, Neuron Model, Parasitic capacitance, Power Consumption
AbstractWhen the large scale Artificial Neural Network is constructed, it is desirable that power consumption of a neuromorphic device is low power consumption. In this paper, we reduce the gate voltage of NMOS as a negative resistance device.As a result, it is clearly shown that power consumption of a neuromorphic device is reduced to about 16.6% when compared with the previous model.

T2-2-6 (Time: 15:10 - 15:30)
TitleAnalysis on the Effect of Heating Coil in Micro-speakers for Smart Car
Author*Hyun-Kab Kim, Hie-Sik Kim (University of Seoul, Republic of Korea), Dalwon Jang, Jongseol Lee, Sei-Jin Jang (KETI, Republic of Korea)
Pagepp. 503 - 505
Keywordvoice coil, transducer, thermal, measurement, micro-speakers
AbstractLoud speakers, which are representative transducers, have non-linearity, thus that requires for speaker researchers to measure loud speakers precisely. In particular, micro-speakers, which are used in potable devices as smartphones, have poor situation for generating sound compared to other sound emitting objects. Thus, more careful measurement is necessary. In this paper, the effect of heating coil through the pre-driver in the micro-speakers with a voice coil is presented. The experimental results related to coil heating is shown, and from the result, we can know coil heating is necessary for precise measurement of micro-speakers.


Session T2-3  [SS7] Information and Communication Technologies for Safe and Secure Life
Time: 13:30 - 15:30 Tuesday, July 12, 2016
Location: Room 3
Chairs: Koichi Gyoda (Shibaura Institute of Technology, Japan), Kazuyuki Miyakita (Niigata University, Japan)

T2-3-1 (Time: 13:30 - 13:50)
TitleProposal of a MANET Broadcast Protocol that Supports Disaster Relief Request
Author*Takuya Kyogashima, Koichi Gyoda (Shibaura Institute of Technology, Japan)
Pagepp. 507 - 510
KeywordMANET, disaster relief, AODV
AbstractIn many existing MANET protocols, it is assumed that the destination IP address is already known. However, when MANET is used in the emergency case such as a disaster occurred, the IP address of the destination may not always be known. In this paper, we propose a MANET protocol that can find a destination IP address by using route request message.

T2-3-2 (Time: 13:50 - 14:10)
TitleA Study on AoIP over MANET for Remote Broadcasting
Author*Omuwa Oyakhire, Koichi Gyoda (Shibaura Institute of Technology, Japan)
Pagepp. 511 - 514
KeywordAoIP, MANET, OLSR, QoS
AbstractAudio over Internet Protocol (AoIP) is one of the emerging protocols that is used for a means of streaming high-quality audio feeds over IP Networks from a remote site to the main studio in radio broadcasting. The AoIP in this study is proposed to be carried over a continuously self-configuring infrastructure-less network of mobile devices refered to as Mobile ad-hoc Networks (MANETs). This study will focus explicitly on the performance of AoIP over MANET with different codecs and proactive protocols. Throughout a simulation study by use of network simulator, the following Quality of Service parameters will be assessed: throughput, delay and packet loss to evaluate the uniqueness of this system

T2-3-3 (Time: 14:10 - 14:30)
TitleStudy on Several Graph Coloring Problems related with Wireless Networks
Author*Shun Matsumoto, Hiroshi Tamura (Chuo University, Japan)
Pagepp. 515 - 518
Keywordwireless network, channel assignment, coloring problem, graph theory
AbstractIn wireless network, each channel assigns to each communication. Then, if the same channel uses at close range, comfortable communication can not be performed by interference. To consider this problem, it is known widely how to replace the channel assignment problem to the coloring problem of the graph. In this paper, we pick up some colorings and a extension of them, examined the minimum number of colors on the graph of tree that the degree is constant.

T2-3-4 (Time: 14:30 - 14:50)
TitleA Consideration on Evaluation of Epidemic Information Sharing by Multiple UAVs
Author*Naoyuki Karasawa, Keisuke Nakano, Kazuyuki Miyakita (Niigata University, Japan), Hiroshi Tamura (Chuo University, Japan)
Pagepp. 519 - 522
Keywordepidemic transmission, UAV, Information Sharing
AbstractIn epidemic communication, mobile nodes carry information by themselves, and send it to other node by direct wireless communication. In this paper, we consider information sharing by using epidemic transmission as a disaster communications, and we use Unmanned Aerial Vehicle (UAV) as a message carrier in a communication between shelters. A UAV travels along the predetermined route and exchanges the information when it arrives at a shelter. After information exchange, the UAV leave the shelter and heads for the next shelter. The UAV must charge its battery in the shelters with charging equipments before their battery is empty. In this paper, we consider the several cases of information sharing between two UAVs at the specific shelter. We evaluate the performance of the information sharing by UAVs by computer simulation.

T2-3-5 (Time: 14:50 - 15:10)
TitlePerformance Evaluation of Information Floating Considering Behavior Changes of Mobile Nodes
AuthorKeisuke Nakano, *Kazuyuki Miyakita (Niigata University, Japan)
Pagepp. 523 - 526
Keywordinformation floating, epidemic transmission, behavior changes, fixed source, theoretical analysis
AbstractEpidemic wireless communication delivers information to destination by spreading information by direct wireless communication between mobile nodes and movement of mobile nodes. To prevent disorderly diffusion of information for a specific local area in epidemic wireless communication, information floating (IF), which restricts mobile nodes to transmit the information only in the transmittable area, has been considered. In this paper, we consider IF for delivery of such information as accident warning and advertisement of a shop. We assume that mobile nodes change direction after receiving the information to avoid the accident site or to approach the shop. We also assume that mobile nodes that pass a fixed source of information (the accident site or the shop) carry the information to the transmittable area. Under these assumptions, we analyze the probability that a node cannot receive information in a transmittable area in the road network where the fixed source of information is surrounded by intersections. We also derive the size of each transmittable area so that this probability is close to desired value.

T2-3-6 (Time: 15:10 - 15:30)
TitleA Pond be Smart for Flood Control
Author*Toshiyuki Moriyama (Fukuoka Institute of Technology, Japan), Katsuhiro Morishita (Kumamoto College, National Institute of Technology, Japan), Shinobu Izumi (Sojo University, Japan), Koji Nishiyama (Kyushu University, Japan)
Pagepp. 527 - 530
Keywordsmart rainwater facilities, rainwater grid, urban river, flood control, distributed multi-purpose dam by citizen
AbstractWe try a second verification to be smart a pond for flood control. To save a rainwater in the pond, we install to monitor the water level, water outlet quantity. And we are planning to dam up the pond by hand using diagnose of heavy rainfall with Self Organizing Map soon.


Session T2-4  Internet Technology & Applications
Time: 13:30 - 15:30 Tuesday, July 12, 2016
Location: Room 4
Chair: Mohammad R. Asharif (University of the Ryukyus, Japan)

T2-4-1 (Time: 13:30 - 13:50)
TitleThai Social Media Alert System for Business
Author*Supatta Viriyavisuthisakul, Parinya Sanguansat, Pisit Charnkeitkong (Panyapiwat Institute of Management, Thailand), Choochart Haruechaiyasak (National Electronics and Computer Technology Center, Thailand)
Pagepp. 531 - 534
KeywordAlert system, Thai social media, Sentiment Classification, Support Vector Machine
AbstractNowadays, social media has exponential growth of information generated by social interaction. In this way, social media has a huge impact on how businesses connect with customers. Pantip is the one of the most popular web board in Thailand in which the customers post their feedbacks about products and services. Among these feedbacks, some of them affects the business in negative way. To solve this problem, business should have a system that can alert their customer’s negative feedbacks on social media. If the business can quickly deal with these negative feedbacks, the problems will be managed more easily. Therefore, many businesses can take advantage of this information for market analysis that can increase their opportunities. In this paper, the social media tools, that can utilized the data from social medias, are surveyed from 2005 to present. Each system has different features, but none of them cannot meet the business requirements, especially when the business cannot deal with the problem immediately because the negative sentiments are incorrectly classified and not alert. This paper proposes the social media alert system based on the business requirements. The machine learning technique is applied here to determine which information should be alerted to business. The system monitors the data in Pantip that relate to the business by keywords. After the data were collected and preprocessed, feature vectors are extracted by Term Frequency-Inversed Document Frequency (TF-IDF) before feeding to Support Vector Machine (SVM). Experimental results show that it can achieved the good accuracy rate and also in terms of the sensitivity and specificity.

T2-4-2 (Time: 13:50 - 14:10)
TitleIntellectual Property Protection using Decentralized Trusted Timestamping Based on the Blockchain
Author*Yuefei Gao, Hajime Nobuhara (University of Tsukuba, Japan)
Pagepp. 535 - 538
Keywordblockchain, timestamp, decentralize, intellectual property
AbstractPresent decentralized trusted timestamping based on the blockchain only provide 40-byte storage for digital files. This does not permit sufficient storage to encode other information. The proposed method expands the storage space to a maximum of N*40 bytes, thereby enabling the storage of additional information (e.g., file name, creator name, and keywords). Experimental results indicate that the proposed method can timestamp a file in an average of 20 min at a possible cost of 0.24 USD. We consider that the proposed method can prove the existence and integrity of a digital file.

T2-4-4 (Time: 14:10 - 14:30)
TitleAdaptive Publish Time and QoS Level over MQTT Protocol
Author*Nuttakit Vatcharatiansakul, Panwit Tuwanut (King Mongkut's Institute of Technology Ladkrabang, Thailand)
Pagepp. 539 - 542
KeywordMQTT, QoS
AbstractSeveral application protocols have been proposed for Internet of Things (IoT) solutions, but the most widely used are the Message Queuing Telemetry Transport (MQTT) and the Constrained Application Protocol (CoAP). To ensure that the messages are transmitted accurately and reliability, both MQTT and CoAP support Quality of Service (QoS). There are three QoS levels in MQTT and four levels in CoAP. However, the characteristic of a medium such as signal attenuation, propagation delays, bit error rates and others affect to this transmission. In this paper, we proposed that QoS level should be adaptively adjusted according to the consideration of link conditions. Moreover, the periodic period times to publish/post message also adaptively adjust. Based on experimental results show that the proposed scheme can ensure message reliability.

T2-4-5 (Time: 14:30 - 14:50)
TitlePatient Health Monitoring System over DDS Protocol
Author*Thanat Maneenut, Panwit Tuwanut (King Mongkut's Institute of Technology Ladkrabang., Thailand)
Pagepp. 543 - 546
KeywordDDS, Patient, real time
AbstractContinuous monitoring of biomedical parameters and/or vital signs of a patient is thus of major importance. The development of the patient monitoring system in real time monitoring is increasingly important. The Internet of things for patient monitoring almost widely uses MQTT and CoAP protocol to collect and send data to hospital or nursing station. However, traffic may be delayed and burst which is not suitable for real-time monitoring. In this paper, we proposed that the patient monitoring system in real time should be use Data Distribution Service (DDS) protocol instead of use MQTT or CoAP, with no broker architecture of DDS, data flow directly from publishers to subscribers. Moreover, the H.264 Scalable Video Coding should be used to encode a video signal, due to the adaptation of bit rate is low complexity, which suitably use in publish-subscribe model.


Session T2-5  Computer Vision (2)
Time: 13:30 - 15:30 Tuesday, July 12, 2016
Location: Room 5
Chair: Ryusuke Miyamoto (Meiji University, Japan)

T2-5-1 (Time: 13:30 - 13:50)
TitleGenerating Lambertian Image by Removing Specular Reflection Component and Difference of Reflectance Factor Using HSV
Author*Naoki Ikeda, Hiroyasu Usami, Yuji Iwahori (Chubu University, Japan), Boonserm Kijsirikul (Chulalongkorn University, Thailand), Kunio Kasugai (Aichi Medical University, Japan)
Pagepp. 547 - 550
KeywordImage Generation, Lambertian Image, Endoscope Image, Uniform Reflectance, HSV
AbstractWhen Lambertian image is assumed to recover the 3-D shape from endoscope image under the assumption of point light source and perspective projection, the image is generally converted to Lambertian image a priori in general. The aim of this paper is to reduce of this calculation cost and to improve the accuracy of results by equalization of reflectance using brightness value V and classification of the hue H. Both of H and V are obtained by HSV conversion from the original RGB image. The effectiveness of the proposed method was evaluated through the experiments using simulation image and actual endoscope image.

T2-5-2 (Time: 13:50 - 14:10)
TitleRealization of Fish Robot Tracking Control Using Position Detecting Algorithm
Author*Kyoo Jae Shin, Yogendra Rao Musunuri (Busan University of Foreign Studies, Republic of Korea)
Pagepp. 551 - 554
Keywordfish robot, control, RF, aquarium robot system, MATLAB and Simulink
AbstractIn this paper, we need to control the motion of the swimming fish robot in order to implement the aquarium fish underwater robot world. And, it implements positional control of the 3axes trajectory path for fish robot. The applied robot has verified the performance though the certificated fields test. It was satisfied the excellent performance such as driving force, durability, and water resistance in experimental results. We can control robot motion that it is to recognize an object by using a video camera without any other sensors inside the fish robot. It is possible to find the position and control the fish robot motion control using RF (Radio Frequency) and controlled through the personal computer. In this paper, we are proposed to realize the control the motion of fish robot tracking control using fish robot position detecting algorithm with MATLAB and Simulink. It was verified by the performance test for the designed aquarium fish robot world.

T2-5-3 (Time: 14:10 - 14:30)
TitleExploring Visual Features of Emotional Images on Social Media
Author*Satoshi Sanjo, Marie Katsurai (Doshisha University, Japan)
Pagepp. 555 - 558
KeywordImage sentiment analysis, Image database, Visualization
AbstractIn this paper, we explore visual characteristics of the emotional images on social media through automatic image generation using a class of Convolutional Neural Networks (CNNs) called deep convolutional generative adversarial networks (DCGANs). Results of experiments present the visual appearance of each emotion, from which we can observe its specific characteristics. Furthermore, based on the trained DCGANs, we investigate the similarity between public image datasets. Our findings are also consistent with sentiment classification across different datasets.

T2-5-4 (Time: 14:30 - 14:50)
TitleAn AR-based Hands-on Study System on Fruits and Vegetables for Preschoolers
Author*Shinobu Tsuchiya, Mitsunori Makino (Chuo University, Japan)
Pagepp. 559 - 562
KeywordAugmented Reality, Education, Hands-on
AbstractIn this paper, an AR-based hands-on study system is proposed, which provides experiences preschoolers about cutting virtual fruits/vegetables. The system consists of a tablet PC with AR system, which displays overlapped images of virtual fruits and vegetables according to cubic AR markers. A preschooler can hold, rotate and cut the virtual foods by manipulate the markers in his/her both hands. Through the intuitive experience, he/she is expected to get interests in structure of the food. Comparing with handling the real foods, the system provides limited but simple, clean and safe environment. The proposed system is evaluated by observation of experiences by preschoolers. From questionnaires to preschool teachers, it is shown that the proposed system can increase the preschoolers’ interests in foods, while some improvements and addition to the features are pointed out.

T2-5-5 (Time: 14:50 - 15:10)
TitleObject Tracking based on Path Similarity of Boosted Decision Trees
Author*Koichi Mitsunari, Jaehoon Yu, Yoshinori Takeuchi, Masaharu Imai (Osaka University, Japan)
Pagepp. 563 - 566
KeywordObject Tracking, Path Similarity, Boosted Decision Trees
AbstractFor general object recognition, detection and tracking are important and complementary components but require different types of features. Therefore, it is difficult to fuse two components. To address the issue, we present a novel tracking method using path similarities between binary decision trees, which are byproducts obtained from detection process. In the experimental result, the proposed method achieved comparable tracking capability to a conventional tracking method using on-line boosting without any extra computation.

T2-5-6 (Time: 15:10 - 15:30)
TitleFeature Learning from Facial Expression Images Using Convolutional Neural Networks
Author*Taiki Nishime (Graduate school of Information Engineering, University of the Ryukyus, Japan), Satoshi Endo, Naruaki Toma, Koji Yamada, Yuhei Akamine (School of Information Engineering, University of the Ryukyus, Japan)
Pagepp. 567 - 570
Keywordfacial expression recognition, convolutional neural networks, deep learning, feature learning
AbstractIn this study, we carried out the facial expression recognition from facial expression dataset using Convolutional Neural Net- works (CNN). In addition, we analyzed intermediate outputs of CNN. As a result, we have obtained emotion recognition score of about 58%; two emotions (Happiness, Surprise) recognition score was about 70%. We also confirmed that specific unit of intermediate layer have learned the feature about Happiness. This paper details these experiments and investigations regarding the influence of CNN learning from facial expression.


Session T2-6  Image Coding & Analysis
Time: 13:30 - 15:30 Tuesday, July 12, 2016
Location: Room 6
Chair: Takeshi Ikenaga (Waseda University, Japan)

T2-6-1 (Time: 13:30 - 13:50)
TitlePre-processing Algorithm using Sub-band Decomposed Multiscale Retinex for Perceptual Video Coding
Author*Kwang Yeon Choi, Ho Hyeong Ryu, Byung Cheol Song (Inha University, Republic of Korea)
Pagepp. 571 - 573
Keywordpre-processing, PVC
AbstractThis paper presents a pre-processing algorithm based on sub-band decomposed multi-scale retinex (SD-MSR) for effective perceptual video coding. First, an input frame is decomposed into multiple sub-bands using SD-MSR. Then, a few high-frequency sub-bands are perceptually suppressed by controlling their gains based on a specific model. Finally, the pre-processed video sequence is encoded via a conventional video encoder. Experimental results show that the proposed algorithm can improve the coding efficiency of H.264 by about 12% on average in comparison with before-use.

T2-6-2 (Time: 13:50 - 14:10)
TitleLossless DWT Transcoder Compatible with lossy DWT for 3D Image
Author*Suvit Poomrittigul (Pathumwan Institute of Technology, Thailand), Masahiro Iwahashi (Nagaoka University of Technology, Japan)
Pagepp. 575 - 578
KeywordTranscoding, DWT, 3D image, JPEG2000, lossless-lossy
AbstractThis paper proposes the lossless discrete wavelet transform (DWT) transcoder compatible to standard lossy coding based on the 9/7 DWT in JPEG2000 (JP2K) for three dimensions (3D) image. The proposed transcoder is applied by re-lifting the integer coding 9/7 DWT and modifying the position of some coefficients. This paper has confirmed that the transcoding between the standard lossless coder 5/3 DWT and the standard lossy decoder 9/7 DWT is not compatible. Since, proposed method presents integer coding 9/7 DWT and more 2 re-lifted integer coding 9/7 DWT to improve the compatibility with standard lossy DWT. The results show that all proposed transcoders are compatible with lossy DWT for 3D image by improving from the PSNR 21.90 up to 57.0 dB.

T2-6-3 (Time: 14:10 - 14:30)
TitleMulti-view 3D CG Image Quality Assessment for Contrast Enhancement Including S-CIELAB Color Space in case the Background Region is Gray Scale
Author*Norifumi Kawabata, Masaru Miyao (Graduate School of Information Science, Nagoya University, Japan)
Pagepp. 579 - 582
KeywordMulti-view 3D Image, Double Stimulus Impairment Scale (DSIS), Luminance, S-CIELAB Color Space, CIEDE2000
AbstractIn this paper, we experimented the subjective evaluation for 3D CG image including the gray scale region with 8 viewpoints parallax barrier method, and we analyzed this result statistically. Next, we measured about the relation between the luminance change and the color difference by using S-CIELAB color space and CIEDE2000. As a result, we obtained knowledge about the relation among the coded image quality, the contrast enhancement, and gray scale.

T2-6-4 (Time: 14:30 - 14:50)
TitleUpper Coding Unit Depth Rate Distortion Cost Comparison Based Fast Intra and Interlayer Inter Prediction Mode Selection for Enhancement Layer in All Intra Spatial SHVC
Author*Pei Liu, Guojing Zhu, Gaoxing Chen, Takeshi Ikenaga (Waseda University, Japan)
Pagepp. 583 - 586
KeywordSHVC, Mode Selection, Enhancement Layer, Coding Unit Depth, Rate Distortion Cost
AbstractSHVC is the scalability extension of the video compression standard high efficiency video coding (HEVC). It encodes the video in different layers, called base layer (BL) and enhancement layer (EL). The additional interlayer inter prediction (ILIP) mode employed in EL increases the complexity. This paper proposes a fast mode selection method for EL in all intra spatial SHVC. The rate distortion cost (RD-cost) difference between intra prediction and ILIP in upper coding unit (CU) depth is calculated as a prejudgment condition to select the prediction mode in lower CU depth. The proposed method can achieve 52.65% and 58.00% encoding time saving for 1.5x and 2x sequences respectively with negligible BD-rate increase compared with SHM-8.0. Compared with related work, this paper can get over 30% time saving increase on average.


Session T3-1  [Tutorial] Unipolar multicarrier modulation for optical wireless communications with intensity modulation and direct detection
Time: 15:50 - 17:50 Tuesday, July 12, 2016
Location: Room 1

T3-1-1
Title(Tutorial) Unipolar Multicarrier Modulation for Optical Wireless Communications with Intensity Modulation and Direct Detection
AuthorPoompat Saengudomlert (Bangkok University, Thailand)
KeywordOptical Wireless Communication, VLC, LED, OFDM
AbstractOptical wireless communications (OWC) has recently gained a lot of attention in the form of visible light communications (VLC). For low-cost system implementations, light emitting diodes (LEDs) are typically used as light sources. Light emitted from LEDs is non-coherent, and hence intensity modulation (IM) is used at LED-based transmitters. Accordingly, power detection or direct detection (DD) is used at receivers. The use of IM/DD requires that data signals are unipolar, i.e., non-negative, making conventional modulation schemes based on bipolar signals not applicable. This tutorial discusses unipolar modulations for OWC with IM/DD. The focus will be on multicarrier modulation, more specifically orthogonal frequency division multiplexing (OFDM), since it is more appropriate than single-carrier modulation for high data rates.


Session T3-2  Analog Circuits (3) and CAD
Time: 15:50 - 17:50 Tuesday, July 12, 2016
Location: Room 2
Chair: Chiranut Sa-ngiamsak (Khon Kaen University, Thailand)

T3-2-1 (Time: 15:50 - 16:10)
TitleNew Approach for Detection and Storage an ESD event by using ReRAM cell
Author*Gi-Doo Lee, Jung-Hoon Chun (College of Information & Communication Engineering, Sungkyunkwan University, Republic of Korea)
Pagepp. 587 - 590
Keywordesd, ReRAM
AbstractThis work proposes a new concept of a circuit that can detect an ESD and save the ESD level by using ReRAM cell. The circuit is configured such that an appropriate voltage level corresponding to the ESD level is applied to both ends of the ReRAM cell, when an ESD event occurs. Through this structure, the resistive state of the ReRAM cell is changed according to the voltage level. The ReRAM cell having a non-volatile characteristic can be maintained the resistance value changed in the ESD event regardless of the power state, and the ESD levels can be identified by the level of the resistance. Using the proposed circuit, a mothod of preventing the soft error caused by sytem level ESD in mobile device is introduced.

T3-2-2 (Time: 16:10 - 16:30)
TitleA Nanopower Full CMOS Sub-Bandgap Voltage Reference
Author*Chang-Bum Park, Shin-Il Lim (Seokyeong University, Republic of Korea)
Pagepp. 591 - 594
KeywordBandgap reference, low voltage, subthreshold, reference circuit, nanowatt
AbstractThis paper describes a sub-1V nanopower full CMOS bandgap reference (BGR) operated in subthreshold region. Complimentary to absolute temperature (CTAT) voltage generator was realized by using two n-MOSFET pair with body bias circuit to make a sufficient amount of CTAT voltage. Proportional to absolute temperature (PTAT) voltage was generated from differential amplifier by using different aspect ratio of input MOSFET pair. The circuits are implemented in 0.18um CMOS process. The simulation results show that the proposed sub-BGR generates a reference output voltage of 358mV, obtaining temperature coefficient of 32 ppm/C in -40°C to 120°C of temperature range. The circuits consume 43nW at 1V supply and operation range is 0.65 to 1.8V.

T3-2-3 (Time: 16:30 - 16:50)
TitleFundamental Analysis of Low Energy Path Routing for Delivery Quadcopters
Author*Shunsuke Negoro, Ittetsu Taniguchi, Hiroyuki Tomiyama (Ritsumeikan University, Japan)
Pagepp. 595 - 596
KeywordQuadcopters, traveling salesman problem (TSP), nearest neighbor method, energy consumption
AbstractThe home delivery service using quadcopters is one of hopeful applications of quadcopters. In this paper, we propose low energy path routing method for delivery quadcopters. Proposed method is based on the nearest neighbor method, and we compared two types of costs functions of low energy path routing problem. In the experiment, we compared two types of proposed methods with optimal solution.

T3-2-4 (Time: 16:50 - 17:10)
TitleVerification of Iterative Algorithms for Electrical-Thermal Analysis of Power Delivery Network
Author*Yasuhiro Nakatani, Tadatoshi Sekine, Hideki Asai (Shizuoka University, Japan)
Pagepp. 597 - 599
Keywordelectrical-thermal co-simulation, iterative algorithm, power delivery network
AbstractIn this paper, electrical-thermal co-simulations of a power delivery network (PDN) are performed by two types of iterative algorithms: Newton's method and an iterative electrical-thermal co-simulation (IETC) method. We show some numerical results of an example PDN to compare the efficiency of the IETC method and that of the conventional Newton's method.

T3-2-5 (Time: 17:10 - 17:30)
TitleValidating Log-normal Distribution of Delay Variability in Near-Threshold Design
Author*Kwangsu Kim, Sangwoo Han, Young Min Park, Minyoung Im, Eui-Young Chung (Yonsei University, Republic of Korea)
Pagepp. 601 - 604
Keywordnear-threshold, log-normal distribution, cell characterization, standard cell library
AbstractIn this work, we perform an in-depth exploration on variability of standard cells operating at near-threshold voltage. Log-normal distribution model displays a very accurate result out of the box,i.e. without further fitting algorithm, although it also yields high error in high-slew conditions and in slew error. We also show that the accuracy of calculated 3 sigma/mean varies with different input slew values. While log-normal modeling is promising, a better fitting methodology is required to overcome its inconsistency in order to adapt to commercial EDA flow.

T3-2-6 (Time: 17:30 - 17:50)
TitleCommunication Aware Compiler for Mesh-Structured Reconfigurable Processors on Single/Multi Chip
Author*Yi Lu, Qinhao Wang, Amir Masoud Gharehbaghi, Masahiro Fujita (The University of Tokyo, Japan)
Pagepp. 605 - 608
KeywordHigh-Level Synthesis, Reconfigurable Architectures, Multi/Many-Core Systems, Integer Linear Programming, Engineering Change Order
AbstractMany-core system performance is still underutilized in many cases. The program optimization on highly parallel systems is hard and usually done manually. The inter-core data transfer delay highly affects the system performance in deep sub-micron age. To overcome these problem, in this paper, we propose an integer linear programming (ILP) based method to analyze and optimize a program running on a mesh-structured processors array. The proposed model includes communication-aware operation binding and mapping. as well as data transfer routing. With this flexible ILP based formulation, optimized binding, mapping and routing is determined for a given program on the target architecture. Our ILP based formulation can also be used for high level ECO while performing high level synthesis.


Session T3-3  Power Electronics & Control
Time: 15:50 - 17:50 Tuesday, July 12, 2016
Location: Room 3
Chair: Lin Min Min Myint (Shinawatra University, Thailand)

T3-3-1 (Time: 15:50 - 16:10)
TitlePower Loss Analysis of Switched-mode Converter Circuits in XMODEL
Author*Yoontaek Lee, Jeongyeol Kwon, Jaeha Kim (Seoul National University, Republic of Korea)
Pagepp. 609 - 612
KeywordSwitched-mode converter, Loss analysis, Modeling, XMODEL
AbstractA way to model a switched-mode converter and estimate its power loss via XMODEL is presented. The model leverages the fast and accurate event-driven simulation capability of XMODEL yet includes the detailed parasitic characteristics of the device components for accurate power loss estimation. The experimental results with a MOSFET-based converters such as a boost converter, and a buck converter demonstrate that the presented model with XMODEL is about 450% faster with <1% error compared to HSPICE simulation.

T3-3-2 (Time: 16:10 - 16:30)
TitleSecure Networked Motion Control Using Tampering Detection Observer
Author*Jin Hoshino, Hitoshi Kojima, Takahiro Funakoshi, Ryusuke Imai, Ryogo Kubo (Keio University, Japan)
Pagepp. 613 - 616
KeywordNetworked Control System, Motion Control, Tampering, Cybersecurity, Time Delay
AbstractThis paper proposes a tampering detection observer (TDO) to achieve safe and secure operation of a networked motion control system. The proposed TDO detects tampering signals as an unexpected disturbance. Simulation results show that the proposed TDO can make it possible to keep stable operation of the networked motion control system with constant time delays even if tampering signals are injected on one of the feedback paths.

T3-3-3 (Time: 16:30 - 16:50)
TitleBurst Data Transmission and Time-Delay Compensation for Energy-Efficient Networked Motion Control
Author*Takahiro Funakoshi, Ryogo Kubo (Keio University, Japan)
Pagepp. 617 - 620
KeywordNetworked Control, Motion Control, Energy Efficiency, Time Delay, Burst Transmission
AbstractThis paper proposes an energy-efficient networked control system (NCS) using burst data transmission and time-delay compensation techniques. The proposed techniques enable energy-efficient operation of the NCS by making its network interfaces enter a low-power sleep mode when no data are exchanged between the controller and plant. Simulations using a networked motion control system show that the proposed techniques can effectively generate idle periods of the network interfaces for energy-efficient operation and achieve stable position control of an electric motor.


Session T3-4  Security
Time: 15:50 - 17:50 Tuesday, July 12, 2016
Location: Room 4
Chair: Tokumi Yokohira (Okayama University, Japan)

T3-4-1 (Time: 15:50 - 16:10)
TitlePublic Batch Auditing Scheme for Cloud Computing
Author*Ren Junn Hwang, Yu Hsuan Yeh (Department of Computer Science and Information Engineering, Tamkang University, Taiwan)
Pagepp. 621 - 624
KeywordCloud Computing, Security, Public Auditing, Batch Public Auditing
AbstractThis paper proposes an efficient batch auditing technique to enhance the security of cloud computing systems. Through this technique, the cloud server computes proofs of audited files on the basis of the proposed auditing tree and sends them to a third-party auditor (TPA). The server uses the auditing tree to temporarily store the proofs during the batch auditing process. If an error message accompanies the batch auditing results, then the cloud server retrieves and sends auditing information from the auditing tree to the TPA without regenerating the auditing information. The TPA quickly and efficiently identifies erroneous files by using secondary information from the cloud server. The communication and computing costs of the proposed technique represent an improvement over similar schemes.

T3-4-2 (Time: 16:10 - 16:30)
TitleLock Access Control and Authorization using Smartphone Based on Bluetooth Communication
Author*Ren Junn Hwang, Bo-Cheng Chen (Department of Computer Science and Information Engineering, Tamkang University, Taiwan), Der-Ren Liao (Advanced Engineering Division, Hua-chuang Automobile Information Technical Center Co., Ltd., Taiwan)
Pagepp. 625 - 628
KeywordDoor Access Control, Bluetooth, Authorization
AbstractMost the mobile devices today are equipped with Bluetooth technology. Features such as low power consumption and medium-to-long transmission range make this technology important for the Internet of Things. This paper proposes an access and authority control scheme based on Bluetooth technology for personal or family usage. In this scheme, a smartphone equipped with Bluetooth acts as a personal trusted device, which can be used as a key device by its owner. The proposed scheme can be applied to the access and authority control of a door, vehicle, or special facility, whose lock is initialized and automatically authenticates the accessing device via Bluetooth without connecting to the server or the Internet. The accessing device is authorized for as long as the user wants to open the lock. The master-phone of the lock can authorize or revoke the disposable or time-limited access right to another phone to open the lock. The proposed scheme provides an efficient and flexible function for managing a varying set of users opening a lock. This scheme is secure as it is resistant to attacks such as replay attack, masquerade attack, modification of messages, man in the middle, and eavesdrop.

T3-4-3 (Time: 16:30 - 16:50)
TitleA Study on EEG evoked by Invisible Visual Stimulation toward Continuous User Verification
Author*Masashi Hattori (Graduate school of Engineering, Tottori University, Japan), Yuki Izumi (Department of Electrical and Electronic Engineering, Tottori University, Japan), Isao Nakanishi (Graduate school of Engineering, Tottori University, Japan)
Pagepp. 629 - 632
KeywordBiometrics, authentication, EEG, invisible visual stimulation
AbstractWe have studied continuous personal authentication using evoked EEGs (Electroencephalograms) by invisible stimuli. Biometric authentication has attracted attention as a means for verifying individuals. In management of users of a system, it is necessary to authenticate them continuously because there is a risk that the users are spoofed by others. Therefore, we try to use EEG evoked by visual stimuli in continuous verification. However, given stimuli can be perceived prevent user’s working. Thus, we attempt to use the evoked potential from stimulation that cannot be perceived (unnoticeable stimulation). The present analysis result shows that the power spectrum of the alpha-band (alpha-band power) which is caused by unnoticeable stimuli may tend to increase more than that caused by stimuli that people can perceive.

T3-4-4 (Time: 16:50 - 17:10)
TitlePeriodic Autocorrelation of A Signed Binary Sequence Additively Generated With Trace Over Odd Characteristic Extension Field
Author*Yasuyuki Nogami (Okayama University, Japan), Satoshi Uehara (The University of Kitakyushu, Japan)
Pagepp. 633 - 636
Keywordsequence, additive, basis, trace, Legendre symbol
AbstractThis paper proposes a new type of signed binary sequence that has an interesting periodic auto-correlation. The sequence is generated by using trace function and Legendre symbol over odd characteristic finite field. As a typical difference from the conventional works, the generation procedure utilizes an additive vector representation and additive vector order.

T3-4-5 (Time: 17:10 - 17:30)
TitleMalware Detection Based on Gray Scale Image and OpCode Feature
Author*Liu Liu, Baosheng Wang, Bo Yu, Yong Tang (National University of Defense Technology, China)
Pagepp. 637 - 640
Keywordgray scale, machine learning, malware
AbstractFirstly, our method transforms the malware binary information into gray scale matrices. Then, we use control flow graph to analyze the operation code. Finally, we classify the malware by combining the gray scale matrices and the opcode features.In the experimental part, we use a variety of classical algorithms to test our method, and the results show that our algorithm has a good advantage.


Session T3-5  Communications (1)
Time: 15:50 - 17:50 Tuesday, July 12, 2016
Location: Room 5
Chair: Masato Saito (University of the Ryukyus, Japan)

T3-5-1 (Time: 15:50 - 16:10)
TitleNovel Compensation Method for Direction of Arrival (DOA) Estimation
Author*Heemang Song, Seunghoon Cho, Kyung-Jin You, Hyun-Chool Shin (Department of Electronic Engineering, Soongsil University, Republic of Korea)
Pagepp. 641 - 644
KeywordAntenna array, Array signal processing, Array beamforming, DOA estimation, Array compensation
AbstractIn this paper, we propose a novel compensation method improving direction of arrival (DOA) and preventing target split tracking. Amplitude and phase mismatching and mutual coupling between radar arrays cause an inaccuracy problem in DOA estimation. By quantifying amplitude and phase distortions for angles, we compensate the distortion. Applying the proposed method to Bartlett, Capon and multiple signal classification (MUSIC) algorithms, we experimentally demonstrate the performance improvement using experimental data from the chamber and real data obtained in actual road.

T3-5-3 (Time: 16:10 - 16:30)
TitleAntenna Scheduling for Energy-Efficient Transmission in Massive MIMO Environments
Author*Geon-Woong Jung, Yong-Suk Byun, Yong-Hwan Lee (Seoul National University, Republic of Korea)
Pagepp. 649 - 652
KeywordMassive MIMO, Energy efficiency, Antenna scheduling, Transmit power allocation
AbstractThe use of massive multiple-input multiple-output (m-MIMO) systems has been considered as one of key technologies in advanced wireless communication systems. However, it may need to employ an energy-efficient transmission technique. The energy efficiency can be improved by using a minimal number of transmit antennas enough to provide desired performance. However, it may require large computational complexity to optimally schedule a set of transmit antennas. In this paper, we consider minimal use of m-MIMO antennas with power control without large computational complexity. We determine the number of transmit antennas and the corresponding antenna subset by exploiting average channel information. We also adjust the transmit power and the spatial multiplexing order to provide desired performance. Finally, we verify the proposed scheme by computer simulation.

T3-5-4 (Time: 16:30 - 16:50)
TitleOn the Effect of Informed Nodes on Learning over Complex Adaptive Networks
AuthorMorteza Farhid, Mousa Shamsi, Mohammad Hossein Sedaaghi (Sahand University of Technology, Iran), Faramarz Alsharif, *Bruno Senzio-Savino, Mohammad Reza Alsharif (University of the Ryukyus, Japan)
Pagepp. 653 - 656
KeywordAdaptive Networks, Diffusion LMS, Mean square deviation, Distributed estimation
AbstractIn this paper, we investigate the impacts of number of informed nodes on the performance of a distributed estimation algorithm, namely adaptive-then-combine diffusion LMS, based on the data with the temporal and spatial independence assumptions. The study covers different network models, including the regular, the small-world and the random networks. We have two scenarios for our simulation. We change the fraction of nodes according to their links densities. The simulation results indicate that the larger proportion of the uninformed nodes (90% in first and up to 50% in second scenarios) in a network causes lower convergence besides improvement in the mean-square-error performance and that acquiring more information is not necessarily better.

T3-5-5 (Time: 16:50 - 17:10)
TitleTail Biting Low Density Parity Check Convolutional Codes over Rain Fade channel
Author*Arisa Wongsriwor (Khon Kaen University, Thailand), Puripong Suthisopapan (Rajamangala University of Technology Isan Khon Kaen, Thailand), Virasit Imtawil (Khon Kaen University, Thailand)
Pagepp. 657 - 660
KeywordKa-band, TB-LDPCCCs
Abstract: Rain fade is one of important effects that can be avoided in Ka-band satellite communications. In order to reduce the effect of rain fade, Tail-Biting low density parity check convolutional codes (TB codes) that have high performance and low complexity is considered. Simulation results show that TB codes with small size of memory can achieve significant coding gain over worst case for rain fade condition. Moreover, the comparison between high performance LDPC and TB codes is also presented. It is found that TB codes with memory of 159 can beat LDPC codes at BER 10-5 .

T3-5-6 (Time: 17:10 - 17:30)
TitleDevelopment of Next-Generation Horticulture Plant based on Wireless-Sensor-Network
Author*Yasushi Shiroma, Ken Kuwae, Shiho Omine, Ryuichi Suwa, Mitsunaga Kinjo, Katsuhiko Shimabukuro, Shiro Tamaki (University of the Ryukyus, Japan)
Pagepp. 661 - 664
KeywordStrawberry Cultivation, Wireless Sensor-Network, Environmental Measurement, Crops environment, Internet of Things
AbstractIn this paper, we developed wireless sensor network and CO2 controller for strawberry farming. Our application to promote growing and increase quality of strawberry in Okinawa prefecture. There is few subtropical regions in Japan. Farmers are often faced with some problems such as environmental measurements. At that time, they solve that problem themself using their experience and expertise. Therefore, we think environmental sensing data represent that and it is an especially important in subtropical regions. We develop an efficient strawberry growing environment monitoring system with automatic measurement technology.


Session T3-6  Multimedia Information Processing
Time: 15:50 - 17:50 Tuesday, July 12, 2016
Location: Room 6
Chair: Hyuk-Jae Lee (Seoul National University, Republic of Korea)

T3-6-1 (Time: 15:50 - 16:10)
TitleAssessing Canine Emotional States Using an Accelerometer
Author*Rina Ouchi (Nara Institute of Science and Technology, Japan), Yuki Maruno (Kyoto Women's University, Japan), Takatomi Kubo (Nara Institute of Science and Technology, Japan), Maki Katayama (Azabu University, Japan), Miho Nagasawa (Jichi Medical University, Japan), Takefumi Kikusui (Azabu University, Japan), Kazushi Ikeda (Nara Institute of Science and Technology, Japan)
Pagepp. 665 - 667
Keyworddog, emotion, accelerometer
AbstractThe estimation of emotional states is important to maximize the performances of rescue dogs. Although such physiological data as heart rate and its variability are effective for estimation tasks, they are difficult to measure in noisy environments. In this paper, we proposed a method for emotional state estimation from acceleration data. The method classified positive, negative, and neutral emotions with 92% accuracy within one subject and 72% over all subjects. These accuracies are high enough for practical use in rescue dogs.

T3-6-2 (Time: 16:10 - 16:30)
TitleHistogram-based Non-Iterative Global Motion Estimation
Author*Kyeongbo Kong, Seungjun Shin, Woo-Jin Song (POSTECH, Republic of Korea)
Pagepp. 669 - 672
KeywordGlobal motion estimation, Motion vector, Robust estimation, Compressed-domain processing, Outlier removal
AbstractWe present a compressed-domain global motion estimation (GME) algorithm that uses a robust estimator. Conventional GME algorithms based on the M-estimator are accurate and powerful, but are slow because they use an iterative process. To speed up the GME, we propose a non-iterative algorithm that exploits an outlier rejection mask and histogram. In simulations the proposed algorithm was about five times faster than the conventional algorithm and had slightly higher estimation accuracy.

T3-6-3 (Time: 16:30 - 16:50)
TitleDigital Market Research using Social Listening Tool for Modern Trade in Thailand
Author*Datchakorn Tancharoen, Supatta Viriyavisuthisakul, Su-amporn Parnsup, Kittipat Panruang (Panyapiwat Institute of Management, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 673 - 676
KeywordDigital Market Research, Social Listening Tool, Modern Trade, Thailand
AbstractModern trade is rapidly developing in Thailand. Market research is an important process for modern trade to know the customer need. Thai people use social media as a communication channel in daily life. They provide the opinions on social media that is useful for market research. Currently, there are some social listening tools which is benefit for digital market research. Thus, this paper presents digital market research using social listening tool for modern trade in Thailand. We apply some social listening tools for some famous brands of supermarket and convenient stores in Thailand. Social listening tools are demonstrated the visualization in terms of features and functions. Many functions are suitable for modern trade in Thailand especially Thai text detection. The experiments show that social listening tools can support digital market research and analyze the modern trade trend. Finally, there are some examples of market research using social listening tools from Thai people comments about modern trade in Thailand.

T3-6-4 (Time: 16:50 - 17:10)
TitleDigital Watermark for Real Musical Instrument Sounds Using Non-negative Matrix Factorization
Author*Harumi Murata (Chukyo University, Japan), Akio Ogihara (Kindai University, Japan)
Pagepp. 677 - 680
Keywordaudio watermarking, non-negative matrix factorization, music theory
AbstractIn this paper, we propose an audio watermarking method using nonnegative matrix factorization (NMF). The amplitude spectrogram of the observed signal is decomposed the basis matrix and the activation matrix by NMF. For embedding watermarks, we use the activation matrix. Onset and offset times are estimated from the coefficients of activation matrix and this interval is defined as duration. The estimated notes are regarded as root notes and watermarks are embedded by operating the activation coefficients of dominant notes.

T3-6-5 (Time: 17:10 - 17:30)
TitleA Chaos-based Watermarking Algorithm for Video Authentication
Author*Jantana Panyavaraporn (Burapha University, Thailand)
Pagepp. 681 - 684
KeywordWatermark, Video, Chaotic Map, Wavelet Domain
AbstractWatermarking can be achieved copyright protection and authentication. This paper proposed a blind digital watermarking algorithm in wavelet domain based on chaotic map. The chaotic map was employed to improve the security and used for embedding position of watermark. The results show the proposed algorithm is more robustness to attacks.

T3-6-6 (Time: 17:30 - 17:50)
TitleThe Discrimination Method for Favorability of Facial Expression Using Convolutional Neural Networks
Author*Masataka Itosu (Graduate School of Information Engineering, University of the Ryukyus, Japan), Satoshi Endo, Nruaki Toma, Koji Yamada, Yuhei Akamine (School of Information Engineering, University of the Ryukyus, Japan)
Pagepp. 685 - 688
KeywordConvolutional Neural Network, facial expression
AbstractThe researches about facial expression and impression pointed out that the beholder feels difference of the impression from subtle difference of the facial expression even as the same person. From these results, I considered that it cam be possible to classify the facial expression by impression. Convolutional Neural Network(CNN) makes top class performance in image recognition field.In this paper, I discuss about the estimation for favorability of smile using CNN as an example of learning the subtle difference of the facial expressions.



Wednesday, July 13, 2016

Session W1-1  Artificial Intelligence (1)
Time: 9:00 - 10:20 Wednesday, July 13, 2016
Location: Room 1
Chairs: Datchakorn Tancharoen (Panyapiwat Institute of Management, Thailand), Rajalida Lipikorn (Chulalongkorn University, Thailand)

W1-1-1 (Time: 9:00 - 9:20)
TitleMachine Learning for Classifying Working State Images Recorded by Digital Tachograph System
Author*Senlin Guan (University of Tsukuba, Japan), Takeshi Shikanai, Morikazu Nakamura (University of the Ryukyus, Japan)
Pagepp. 689 - 692
Keywordmachine learning, support vector machine, time-series image, Digital Tachograph System, sugarcane
AbstractMachine learning is a powerful modelling and prediction tool for data analysis and decision-making in agriculture production, especially for the cases of dealing with large volume of data in diverse formats. In this paper, we present a case study of applying machine learning to classify the working states of harvesting sugarcane based on the time-series data, which is recorded by a digital tachograph system mounted on a small sugarcane harvester. The study aims at constructing a model and training it applicable to automatically learn from the time-series images and classify the images into different working states. Three machine learning models are implemented to evaluate the best accuracy for classification and the optimum parameters for the model. The result indicates that using machine learning is an effective way to distinguish the working states, and the average F1_score reaches 0.970 when recognizing the cutting state. The classification by the Support Vector Machine (SVM) model with Radial Basis Function (RBF) kernel gains higher accuracy than by that with linear kernel and by K-Nearest Neighbors (KNN).

W1-1-2 (Time: 9:20 - 9:40)
TitleApplying Purchase Porter System to Actual Shopping Mall
Author*Hiroki Nakazaki, Katsumi Harashima (Osaka Institute of Technology, Japan)
Pagepp. 693 - 696
Keywordautonomous behavior, own work area, bucket brigade method
AbstractWe propose “Purchases Robot Delivery System” for shoppers in this paper. Each robot performs autonomous behavior and delivers purchases to the exit that shoppers use by the time shoppers go home. Delivery robots acting in only own work area cooperate each other and deliver all purchases autonomously by a bucket brigade method because delivery requests occur from many stores in various timing.

W1-1-3 (Time: 9:40 - 10:00)
TitleThe Association and Prediction of Index of the Stock Exchange of Thailand (SET) between the Index of Foreign Exchange by using Data Mining Techniques
Author*Sayan Tepdang, Tortanut Choosang, Suvijak Sothanakul (Rajamangala University of Technology Tawan-Ok Chakrabongse Bhuvanarth Campus, Thailand)
Pagepp. 697 - 700
KeywordArtificial Intelligence, Data Mining, Ensemble, Neural Networks, FP-Growth Algorithm
AbstractThis paper presents the association and prediction of the Stock Exchange of Thailand (SET) by using data the index of foreign exchange from Bloomberg news agency during September 8, 2014 to September 7, 2015. The first stage of the research is to find the association of information which has an increase or decrease of the Stock Exchange of Thailand with the composite index of Shanghai, China, Nikkei 225 Stock index of Japan, the index of the New York Stock Exchange, USA and the index of the London Stock Exchange, UK. Are they relevant after the close of trading on the stock market every day? The method is FP-Growth Algorithm. The results of the research showed that the Support value is greater than 10% of data. This can create association rules related to Stock Exchange of Thailand index which has a total of 56 law relationships; for example: when the exchange composite, Stock exchange Nikkei 225, New York Stock Exchange and the London Stock Exchange are at the close of trading and show the rise of stock market indices all four. This indicates that the close of trading on the Stock Exchange of Thailand will have to increase the index. The confidence is up to 70% and the lift is at 1.443. At the second stage of the research, after association rules related Stock Exchange of Thailand, the selected rule related with the confidence more than 60%. To create a model for prediction the index of SET, by means of Ensemble consisting of three ways: Decision Tree, K-Nearest Neighbors (K-NN) and Neural Networks. The results of the research showed that modeling of the London Stock Exchange has a value of up to 63.48% in Accuracy, followed by the two equal models: Modeling relationship rules of the Shanghai Stock Exchange and the London Stock Exchange composite. The second model is law relationship Stock Exchange Nikkei 225 and the London Stock Exchange with a value of Accuracy equal to 63.38% and the minimum is to create a model law relationship Stock Exchange composite of Shanghai and New York Stock Exchange is worth at 58.43% of Accuracy.

W1-1-4 (Time: 10:00 - 10:20)
TitleA New Double Adaptation Algorithm for Acoustic Noise Control
AuthorMohammad Reza Asharif, *Linnan Zhang (University of the Ryukyus, Japan), Mahdi Khosravy (University of Information Science and Technology, Macedonia, Republic of), Bruno Barzellato, Faramarz Asharif (University of the Ryukyus, Japan), Katsumi Yamashita (Osaka Prefecture University, Japan)
Pagepp. 701 - 704
KeywordAdaptive Filtering, LMS, FXLMS, ANC, Secondary path
AbstractNowadays, Acoustic Noise Cancellation (ANC) is widely used for different applications. One of them is cancellation of noise generated by fans, such as the ones found in air conditioning, computers, ventilation systems, etc.. Normally, fan noise is in the form of an unwanted audio signal. Here, by processing the fan noise record through the microphones, it is adaptively estimated and cancelled. A novel adaptive filter for estimation of two paths at the same time is presented as it is called double FXLMS.


Session W1-2  Intelligent Transportation Systems
Time: 9:00 - 10:20 Wednesday, July 13, 2016
Location: Room 2
Chair: Itaru Nagayama (University of the Ryukyus, Japan)

W1-2-1 (Time: 9:00 - 9:20)
TitleAn Automatic Traffic Light Control System using Microcontroller and Fuzzy Logic Algorithm
Author*Ekkachai Payapwattanawong, Panwit Tuwanut (King Mongkut's Institute of Technology Ladkrabang, Thailand)
Pagepp. 705 - 708
KeywordTraffic Light, Transportation, Road, Fuzzy Logic, Microcontroller
AbstractThe traffic congestion is the problem in urban areas. Traffic management and optimization can help provide better life quality to people. The advanced technology can help to analyze real-time traffic conditions and identify changes in the traffic flow. In this paper, the traffic control system will be proposed with a microcontroller and the Fuzzy Logic Algorithm. This system will be installed at an intersection to help reduce the delay during the traffic light changing to optimize the traffic flow.

W1-2-2 (Time: 9:20 - 9:40)
TitleAdvanced Traffic Prediction System by Socio-Technical Sensor Fusion using Machine Learning
Author*Swe Swe Aung, Shiro Tamaki, Itaru Nagayama (University of the Ryukyus, Japan)
Pagepp. 709 - 712
KeywordTraffic Prediction Server
AbstractNowadays, as traffic jam is an everyday facing problem in the developed and developing countries, monitoring, predicting and detection current traffic condition systems are playing an important role in research fields. The paper mainly focus on predicting traffic condition based on multiple points of view such as the data from road side camera, weather condition, weekday or weekend, rush hour time and special day. The responsibility of Traffic Prediction Server is to the integrate history data and the current input data by applying Naive Bayes Classifier, using m-estimate of probabilities.

W1-2-3 (Time: 9:40 - 10:00)
TitleNew Optical-Wireless CSK-MPPM System with Modified Prime Sequence Code
Author*Masayuki Ishikawa, Hiromasa Habuchi, Atsuhiro Takahashi (Ibaraki University, Japan)
Pagepp. 713 - 716
Keywordoptical hierarchical modulation method, code shift keying, multi-pulse pulse position modulation, GMPSC, ITS
AbstractIn this paper, a new hierarchical CSK(code shift keying)-MPPM(multi-pulse pulse modulation) system is proposed. The system improves reliability of intelligent transport systems. Although it is difficult to set the optinum threshold value, many conventional hierarchical systems use the threshold detector. The proposed system demodulates both of signals by comparative decision. In this paper, the bit error rate of the proposed system and the conventional MPPM-CNK(code number keying) is evaluated. It is shown that CSK-MPPM outperformes MPPM-CNK under the complete synchronization.

W1-2-4 (Time: 10:00 - 10:20)
TitleMultiple Destination Multiple Route Recommendation for Evacuation in Emergency Situation
Author*Chayanon Sub-r-pa, Goutam Chakraborty (Iwate Prefectural University, Japan)
Pagepp. 717 - 720
KeywordRoute Recommendation, Road network
AbstractIn road network multiple destinations route search- ing is common when users are searching for service instead of destination. Navigation system will present list of destina- tion, then user will choose the specific destination. For more constrained situation such as request for parking-lot near to festival ground or request for safety shelter for disaster situa- tion. Navigation system will show list of destinations in the area, then user will choose specific destination by themselves. In general, user will choose the closest destination and short- est route. If there are many users searching the same service simultaneously, the closest destination and the shortest route will get traffic congestion quickly. Vehicles will face traffic congestion when follow the route, and may not get service at destination, because each destination has limited resource. It would be better to choose different route or different destina- tion from beginning of travel. In this paper, we proposed road network routing algorithm considering multiple destinations, multiple routes, and resources at destinations are limited.


Session W1-3  Antenna & Wave Propagation
Time: 9:00 - 10:20 Wednesday, July 13, 2016
Location: Room 3
Chair: Vitawat Sittakul (King Mongkut's University of Technology North Bangkok, Thailand)

W1-3-1 (Time: 9:00 - 9:20)
TitleFade Slope Comparison between Thailand and Japan for Ka band Rain Attenuation
Author*Peeramed Chodkaveekityada, Hajime Fukuchi (Tokyo Metropolitan University, Japan)
Pagepp. 721 - 724
KeywordFade slope, Frequency scaling, Tropical attenuation, Temperate attenuation
AbstractFade slope can be expressed the localized rain behavior using attenuation beacon data. This paper used Thaicom 2 and 3 beacon received in Thailand to represent rain occurrence in tropical areas and CS beacon received in Japan represents for temperate areas. Thaicom 2, 3 and CS beacon operated different frequency, the scaling formula from the International Telecommunications Union (ITU) R P.618-12 [1] standard is used for the comparison. Thaicom beacon frequencies from 12.57 (Thaicom 2) and 12.59 (Thaicom3) GHz in the Ku band are scaling up to the similar of CS beacon at 19.45 GHz. Moreover, we derived the indicators such as standard deviation and skewness obtained from fade slope to confirm rain attenuation activities in both regions.

W1-3-2 (Time: 9:20 - 9:40)
TitleA Wideband Thin Film Slot Array Antenna using Mylar Polyester
Author*Chawalit Rakluea, Paitoon Rakluea (Department of Electronic and Telecommunication Engineering, Faculty of Engineering, Rajamangala University of Technology Thanyaburi, Thailand)
Pagepp. 725 - 728
KeywordSlot Array Antenna, Mylar Polyester, Wideband
AbstractThis paper presents a Slot Array Antenna (SAA) which is fabricated on Thin Film Mylar Polyester (TFMP) substrate. The TFMP is low cost and also flexible material which can bend on curved surface and used for on-body communication system. The prototype antenna consists of microstrip transmission line and periodic slot arrays. For the measurement results, SAA shows dual resonant bahaviors at 2000 MHz and 2200 MHz with similar bidirectional radiation. The bandwidth is 520 MHz or 24.76%. In addition, the average measured gain is around 10.1 dBi at its operation bandwidth. Finally, SAA using TFMP can achieve wideband, high gain and conformal antenna.

W1-3-3 (Time: 9:40 - 10:00)
TitleTriple Band-Notched Thin-Film Ultra-Wideband Antenna fed by CPW
AuthorPaitoon Rakluea (Department of Electronic and Telecommunication Engineering, Faculty of Engineering, Rajamangala University of Technology Thanyaburi, Thailand), *Maleeya Tangjitjetsada (Department of Electronic and Telecommunication Engineering, Faculty of Engineering and Architecture, Rajamangala University of Technology Suvarnabhumi, Thailand)
Pagepp. 729 - 732
KeywordThin-film UWB antenna, Triple band-notched, CPW-fed
AbstractThe thin-film Ultra-Wideband (UWB) antenna fed by Coplanar Waveguide (CPW) with triple band-notched characteristics is presented in this paper. This proposed antenna has concise dimensions of 34.5×27.6 mm2 and is fabricated on flexible thin-film Mylar Polyester. In addition, the thickness and dielectric constant of polyester is 0.3 mm. and 3.2, respectively. The prototype antenna consists of fork-like tuning stub fed by CPW and rectangular slot. Moreover, two strait slots and U-shape slot is added on the top ground plane to accomplish triple band-notched at frequency range of 3.3 – 3.7 GHz (WiMAX), 5.1 – 5.8 GHz (WLAN), and 7.25 – 7.75 GHz (X-Band downlink). For measurements of antenna, the operating frequncy is 2.8 – 13.5 GHz which has three notched-frequency at 3.5, 5.5, and 7.5 GHz. Furthermore, the measured gain is average 3.6 dBi entire bandwidth except notched frequencies which decrease extremely lower than 0 dBi. As the results of measurement, the prototype antenna can appropriately attain triple notched-band for WLAN, WiMAX, and X-Band Communication and is also properly used in UWB equiptment and system.


Session W1-4  Optical Communications and Components (1)
Time: 9:00 - 10:20 Wednesday, July 13, 2016
Location: Room 4
Chair: Cheon Won Choi (Dankook University, Republic of Korea)

W1-4-1 (Time: 9:00 - 9:20)
TitleBroadband Data Transmission over Radio over Fiber
Author*Vitawat Sittakul, Sirivat Hongthong (King Mongkut’s University of Technology North Bangkok, Thailand)
Pagepp. 733 - 734
KeywordRoF, optical, communication, fiber
AbstractThe demand for broadband data transmission has rapidly increased due to the need of enormous data usage. New applications require high bandwidth for transmitting the data such as voice and video over 3G/4G mobile communication, wireless local area network and Gigabit Ethernet. To cope with this situation, broadband data transmission over Radio over Fiber (RoF) has been chosen. Here, a RoF link is created using a 1550nm laser diode (LD) and a photodiode (PD). By carefully selecting the LD and PD, broadband data transmission of ~1.5 GHz can be possibly achieved, measured by the Vector Network Analyzer (VNA). Finally, the modulated data of BPSK, and QAM modulation at 900MHz carrier frequency from the output port of first Universal Software Radio Peripheral (USRP) have been generated and transmitted over the RoF link to the input port of second USRP to measure the link performance such as Bit-Error-Rate (BER), including the noise figure calcualtion of the link.

W1-4-2 (Time: 9:20 - 9:40)
TitleNovel Parallel Optical Turbo-code Communication System using RGB LEDs for Partial Erasure Channel
Author*Ran Sun, Hiromasa Habuchi (Ibaraki University, Japan), Yusuke Kozawa (Tokyo University of Science, Japan)
Pagepp. 735 - 738
Keywordoptical wireless communication, turbo-code, partial erasure channel, RGB parallel transfer
AbstractIn optical parallel communications using RGB LEDs, the different attenuation rates of each colors cause the partial erasure channel, which makes the performance of communications down seriously. In this paper, a novel parallel optical Turbo-code communication system using RGB LEDs is proposed and investigated. It is assumed to be effective to solve the partial erasure channel transfer problem by re-arranging the Turbo-code parity-bit streams. Binary Pulse Position Modulation (BPPM) is used as the modulation technique. Moreover, for showing the effectiveness of the proposed system, the bit error rate (BER) performance of the proposed system is analyzed by computer simulation in Addict White Gaussian Noise (AWGN) channel and Optical Wireless Channel (OWC) taking background noise and scintillation in account, respectively.

W1-4-3 (Time: 9:40 - 10:00)
TitleFast-Recovery p-Cycle Protection in WDM Networks Using Straddling Segments
Author*Quynh Le (Chulalongkorn University, Thailand), Poompat Saengudomlert (Bangkok University, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 739 - 742
Keywordprotection and restoration, flow p-cycle, WDM, survivability, p-cycle
AbstractThe path-segment-protecting p-cycle (or “flow p-cycle”) contributes more saving on spare capacity than conventional link-protecting p-cycle. However, this technique fundamentally requires signaling at all nodes on the restoration paths. This paper presents a new approach on routing the light-paths of node pairs in order to boost the amount of traffic traversing straddling segments in opposite directions so that both end nodes of each segment can detect any link failure on the segment and perform protection switching around a p-cycle.

W1-4-4 (Time: 10:00 - 10:20)
TitleExtended p-Cycle for Path-Segment Protection in Mixed Line Rate WDM Network with Limited Transmission Reaches
Author*Quynh Le, Puttipong Sornplaeng (Chulalongkorn University, Thailand), Poompat Saengudomlert (Bangkok University, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 743 - 746
Keywordp-cycle, path-segment, protection, mixed line rate, transmission reaches
AbstractWe consider the problem of routing and rate assignment with extended p-cycle for path-segment protection (EPPSP) for mixed line rate (MLR) networks that are limited by transmission reaches. We perform a comparative study between EPPSP and conventional p-cycle approaches. Numerical results demonstrate that the EPPSP approach yields lower costs and is more efficient for network protection than conventional p-cycle protection.


Session W1-5  Communications (2)
Time: 9:00 - 10:20 Wednesday, July 13, 2016
Location: Room 5
Chair: Shiro Tamaki (University of the Ryukyus, Japan)

W1-5-1 (Time: 9:00 - 9:20)
TitleKeyword-Based Search Feature Using Keyword-Based Breadcrumbs and Independent Search and Merge for Content-Oriented Networks
Author*Kevin Pognart, Yosuke Tanigawa, Hideki Tode (Osaka Prefecture University, Japan)
Pagepp. 747 - 750
KeywordContent-Oriented Network, search, keyword, cache, Information Centric Network
AbstractContent-Oriented Networks (CON) are another approach to the current Internet to optimize content sharing. Routing focuses on contents themselves instead of locations. In CON, content naming scheme is essential but it can make content searches very difficult from users’ point of view. Hence we propose to integrate a keyword-based search feature to CON. This feature is based on two existing systems: Keyword-based Breadcrumbs for having information about hot contents, and Independent Search and Merge for distributing and managing all content information over the network. This system assures search efficiency and it can rank results by knowing hot contents.

W1-5-2 (Time: 9:20 - 9:40)
TitleDesign of Equalizer based on Bernstein Polynomials under Echo Pairs
Author*Virote Pirajnanchai, Paitoon Rakluea (Department of Electronic and Telecommunication Engineering, Faculty of Engineering, Rajamangala University of Technology Thanyaburi, Thailand)
Pagepp. 751 - 754
KeywordEqualizer, Bernstein, Echo Pairs
AbstractThis paper presents design of equalizer based under echo pairs and amplitude distortion. The equalizer was used with the fourth order bernstein polynomials. The advantage of this design is to get smooth magnitude response and zero phase response. It was found possible to eliminate echo pairs and amplitude distortions. The modulated sine-squared pulse test signal was used for test the performance of proposed equalizer. This research attempted to implement the equalizer with the MATLAB software. The simulation and experimental results are in good agreement. It is likely that application of equalizer to broadcast television systems.

W1-5-3 (Time: 9:40 - 10:00)
TitleActive Queue Management Using Remote Congestion Controller with Model-Free Butterfly-Shaped Perfect Delay Compensator
Author*Ryosuke Hotchi, Hosho Chibana, Ryogo Kubo (Keio University, Japan)
Pagepp. 755 - 758
KeywordAQM, TCP, Time Delay, Control Theory, QoS
AbstractThis paper proposes a remote congestion controller with the butterfly-shaped perfect delay compensator (PDC) for time-delay compensation in active queue management (AQM) supporting transmission control protocol (TCP) flows. The butterfly-shaped PDC does not need any time-delay model of TCP/AQM network. Simulations show that the proposed controller with the butterfly-shaped PDC effectively stabilizes the TCP/AQM network even if the system includes constant or time-varying delays.

W1-5-4 (Time: 10:00 - 10:20)
TitleConstruction of Column-Weight-Two Irregular LDPC Codes with a Predetermined High Girth
Author*Gan Srirutchataboom, Nut Tantibut (Chulalongkorn University, Thailand), Piya Kovintavewat (Nakhon Pathom Rajabhat University, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand), Kritsada Mamat (Kasetsart University, Thailand), Watid Phakphisut (Bangkok Suvarnabhumi University, Thailand)
Pagepp. 759 - 762
KeywordLDPC
AbstractA girth is one of the parameters that can determine the performance of low-density parity-check (LDPC) codes. Because a full Tanner graph of a parity- check matrix, H, can be generated from a small graph, we propose a novel algorithm for constructing irreg- ular column-weight-two LDPC codes with a predeter- mined large girth. The proposed algorithm carefully add edges and vertices in the graph until it fully expands to represent the required H matrix. Simulation results show that the proposed algorithm can provide the H matrix with a very high girth if compared to a well- known progressive-edge growth (PEG) algorithm, and also yields a good bit-error rate performance.


Session W1-6  Audio / Speech Signal Processing (1)
Time: 9:00 - 10:20 Wednesday, July 13, 2016
Location: Room 6
Chairs: Kenji Suyama (Tokyo Denki University, Japan), Tomohisa Wada (University of the Ryukyus, Japan)

W1-6-1 (Time: 9:00 - 9:20)
TitleSpeech Recognition using MFCC with Time Varying LPC for Similar Pronunciation Phrases
Author*George Mufungulwa, Alia Asheralieva, Hiroshi Tsutsui, Yoshikazu Miyanaga (Hokkaido University, Japan)
Pagepp. 763 - 766
KeywordMFCC, time-varying LPC, HMM, Speech Recognition, Similar Pronunciation
AbstractNoises pause a serious challenge in speech recognition systems. Both multiplicative and additive noises affect the performance of automatic speech recognition (ASR) systems. In this paper, a method which makes use of fast Fourier transform (FFT) based Mel-frequency cepstral coefficients (MFCCs) with time-varying linear predictive coding (TVLPC) is proposed. Evaluation results of similar pronunciation phrases and 142 words recognition experiments demonstrate that the proposed approach achieves better results at 10dB SNR than the conventional approach.

W1-6-2 (Time: 9:20 - 9:40)
TitleDesign of IIR Filters Using PSO with Improved Intensification and Diversification Ability
Author*Kenzo Yamamoto, Kenji Suyama (Tokyo Denki University, Japan)
Pagepp. 767 - 770
KeywordIIR filter, PSO, Stochastic optimization
AbstractIn this paper, a design method of IIR filters using PSO is studied. In our previous work, an avoidance method of local minimum stagnation has been proposed, in which a penalty is added to an objective function. In this study, a novel method which repeats a diversification and an intensification alternatively is proposed for a single-swarm PSO, in which a penalty range is fixed to a small value. Several design examples are shown to present the effectiveness of the proposed method.

W1-6-3 (Time: 9:40 - 10:00)
TitleDesign of Constrained IIR Filters Using PSO
Author*Haruna Aimi, Kenji Suyama (Tokyo Denki University, Japan)
Pagepp. 771 - 774
KeywordIIR filter, PSO
AbstractIn this study, IIR filters having a null frequency in a stopband are designed using PSO. A new penalty function is introduced to an objective function in addition to a conventional penalty function which ensures a stability of IIR filters. Then, local minimums are brought to the objective function by adding such a constraint. Therefore, it is important to avoid a local minimum stagnation of PSO. In the proposed method, a particle reallocation method is applied when the stagnation has occurred. The effectiveness of the method is verified through several design examples.

W1-6-4 (Time: 10:00 - 10:20)
TitleMultiple Sound Source Tracking via An Avoidance of Spatial Aliasing
Author*Katsuya Nakazawa, Kenji Suyama (Tokyo Denki University, Japan)
Pagepp. 775 - 778
Keywordmicrophone array, spatial aliasing, delayed sum array method
AbstractIn this paper, a method of multiple sound source tracking using microphone array is studied. MUSIC is known as a high spatial resolution method. However, it requires high computation cost. For reducing the cost, PAST-IPLS based on MUSIC has been proposed. Although an evaluation function of IPLS tends to indicate the high spatial resolution by expanding a microphone width, several local minimums tend to appear.Therefore, we propose an avoidance method of spatial aliasing by using a multiplication of evaluation functions of whole frequency bands.


Session W2-1  Artificial Intelligence (2)
Time: 10:30 - 11:50 Wednesday, July 13, 2016
Location: Room 1
Chair: Dongshik Kang (University of the Ryukyus, Japan)

W2-1-1 (Time: 10:30 - 10:50)
TitleBrain Wave Pattern Classification from Virtual Training Environment by Self-Organizing Maps
Author*Bruno Senzio-Savino, Mohammad Reza Alsharif (University of the Ryukyus, Japan), Carlos E. Gutierrez (Okinawa Institute of Science and Technology, Japan), Christian Penaloza (Advanced Telecommunications Research Institute International, Japan), Katsumi Yamashita (Osaka Prefecture University, Japan), Faramarz Alsharif, Mahdi Khosravy (University of the Ryukyus, Japan), Mousa Shamsi (Sahand University of Technology, Iran)
Pagepp. 779 - 782
KeywordBMI, SOM, Graphical User Interface, Classification, EEG
AbstractThe main purpose of this research is to perform and analyze the performance of a simple brain wave password signal pattern feature classifier by Self Organizing Maps (SOM) and the processing time dedicated to build it. These signals are composed by attention/meditation signal patterns obtained with the aid of a virtual training environment. The signal treatment for the obtained data is explained and the output results from 10-fold cross validation are presented for different feature vector classification by a simple Kohonen layer SOM. The best result represents a 60% average recognition rate for an average processing time of about 132 seconds.

W2-1-2 (Time: 10:50 - 11:10)
TitleDeriving ECG from EASI Electrodes via Machine Learning
Author*Piroon Kaewfoongrungsi, Daranee Hormdee (Khon Kaen University, Thailand)
Pagepp. 783 - 786
Keyword12-lead System, EASI Electrodes, Dower Method, SVR, Machine Learning
AbstractThe 12-lead Electrocardiogram (ECG) is the standard clinical method of heart disease diagnose. Measuring all 12 leads is often impractical. In 1988, Gordon Dower has introduced an EASI-lead ECG System. In order to gain all 12-lead ECG back from this EASI-lead system, Dower's equation was proposed then. Ever since various attempts have been explored to improve the synthesis accuracy, mostly via linear regression. This paper presents how machine learning was used to find a set of transfer function for deriving the 12-lead ECG from EASI-lead system. The experiments were conducted to compare the results those of Support Vector Regression (SVR), Artificial Neural Networks (ANNs) against those of Dower's method. The results have shown that the best performance amongst those methods with the less RMSE error values for all signals with the standard 12-lead ECG was obtained by SVR, followed ANNs and Dower's equation, respectively.

W2-1-3 (Time: 11:10 - 11:30)
TitleTractor Oil Pump Fault Diagnosis by Pseudo-spectrum Analysis of Vehicle Sound Records
AuthorSaurabh Gupta (John Deere India PVT. LTD., India), Mahdi Khosravy, Neeraj Gupta, Bhupendra Nath Tiwari (University of Information Science and Technology, Macedonia, Republic of), *Bruno Senzio-Savino, Faramarz Asharif, Mohammad Reza Asharif (University of the Ryukyus, Japan)
Pagepp. 787 - 790
KeywordVehicle part defect, Tractor oil pump, pesudo-spectrum
AbstractA pseudo-spectrum analysis approach for tractor oil pump defect detection through sound records in a simple mobile application is presented. The sound of tractor engine is recorded in different conditions of oil pump is contamination to dust. By analyzing the sound tracks the dust pollution level is determined and chocking of oil pump is efficiently detected.

W2-1-4 (Time: 11:30 - 11:50)
TitleComparison of ANN and SVM for Prediction of Biochemical Oxygen Demand in Chaophraya River
Author*Weeris Treeratanajaru (Machine Intelligence and Multimedia Information Technology Lab, Department of Mathematics and Computer Science, Faculty of Science, Chulalongkorn University, Thailand), Supawin Watcharamul (Department of Environmental Science, Faculty of Science, Chulalongkorn University, Thailand), Rajalida Lipikorn (Machine Intelligence and Multimedia Information Technology Lab, Department of Mathematics and Computer Science, Faculty of Science, Chulalongkorn University, Thailand)
Pagepp. 791 - 793
KeywordANN, SVM, Water quality model
AbstractArtificial Neural Network (ANN) and Support Vector Machine (SVM) models are used increasingly to predict, monitor and forecast water quality. In this paper, two methods were implemented to predict biochemical oxygen demand (BOD) of Chaophraya River, Thailand using a set of simple measurable surface water quality variables including water temperature, dissolved oxygen (DO), electrical conductivity (EC), pH, nitrate, ammonia, total phosphate (TP), monitoring time, and monitoring location as input variables. The data set consists of 1248 water samples represent 18 different monitoring stations along the Chaophaya, which has been monitored for 17 years. The associated parameters for optimum ANN and SVM model were obtained using grid search technique. The ANN and SVM models can predict BOD in training and testing data sets with reasonably high correlation. The overall results showed that both models could be used as one of the fast, reliable and cost-effective methods for predicting BOD in environments.


Session W2-2  Medical Electronics & Circuits
Time: 10:30 - 11:50 Wednesday, July 13, 2016
Location: Room 2
Chair: Kazuhiro Shouno (University of Tsukuba, Japan)

W2-2-1 (Time: 10:30 - 10:50)
TitleAll-Digital and Low-Power Reference Clock Generator for Biotelemetry Applications
AuthorDuo Sheng, Chia-Lin Wu, *Yu-Chan Hung, Yi-Shang Wang (Fu Jen Catholic University, Taiwan)
Pagepp. 795 - 798
KeywordWBAN, Clocking, Low Power
AbstractIn this paper, an all-digital and low-power reference clock generator with cell-based design for biotelemetry applications is presented. The proposed clock generator employs a cascade-stage structure to achieve high resolution and wide range at the same time. Besides, based on the proposed Schmitt-trigger-based delay cell (STDC), the proposed clock generator not only can provide high resolution, but also can generate low frequency clock signal with low power consumption and low circuit complexity as compared with conventional approaches. Simulation results show that the operation frequency range is from 7.9MHz to 120.8MHz, and the power consumption can be improved to 54μW (@7.9MHz) with 10.5ps resolution. In addition, the proposed clock generator can be implemented with standard cells, making it easily portable to different processes and very suitable for biomedical System-On-Chip (SoC) applications.

W2-2-2 (Time: 10:50 - 11:10)
TitleBurst Suppression Pattern Recognition Using Time-Frequency Analysis
Author*Jaeyun Lee, Woo-Jin Song (POSTECH, Republic of Korea), Hyun-Chool Shin (Soongsil University, Republic of Korea)
Pagepp. 799 - 802
KeywordEEG, Burst suppression, Maximum likelihood estimation, Time-frequency analysis
AbstractWe developed a method to recognize EEG burst suppression in the joint time-frequency domain. We obtained the feature used in the proposed method from the joint use of the time and frequency domains, and we estimated the decision as to whether the measured EEG was a burst segment or suppression segment by a maximum likelihood estimation. We evaluated the performance of the proposed method in terms of its accordance to the visual score and estimation of the burst suppression ratio. The accuracy was higher than the sole use of the time domain, as well as conventional methods conducted in the time domain. In addition, probabilistic modeling provided a more simplified optimization than conventional methods. Quantification of burst suppression necessitated precise burst suppression segmentation with an easy optimization; therefore, the excellent discrimination of burst suppression by the proposed method appears beneficial.

W2-2-3 (Time: 11:10 - 11:30)
TitleTwo Coexisting Two-dimensional Tori Generated in a Three-coupled Delayed Logistic Map
Author*Naohiko Inaba (Meiji University, Japan), Munehisa Sekikawa (Utsunomiya University, Japan), Daiki Ogusu, Tetsuro Endo (Meiji University, Japan)
Pagepp. 803 - 806
KeywordQuasi-periodic bifurcations
AbstractQuasi-periodic bifurcations have attracted considerable attention in recent years. In this study, we discuss two coexisting two-dimensional tori in an Arnol'd tongue generated in a three coupled delayed logistic map. The two coexisting two-dimensional tori comprise 93 invariant closed curves. One of two-dimensional tori disappear by a quasi-periodic saddle-node bifurcation, and the other two-dimensional torus bifurcates to a three-dimensional torus via a quasi-periodic saddle-node cycle bifurcation. The generation of the three-dimensional torus is confirmed by observing the attractor on a double Poincare section.


Session W2-4  Optical Communications and Components (2)
Time: 10:30 - 11:50 Wednesday, July 13, 2016
Location: Room 4
Chair: Poompat Saengudomlert (Bangkok University, Thailand)

W2-4-1 (Time: 10:30 - 10:50)
TitleProposal of a Variable N-CSK with MPSC for an Intellectual Illumination Light Data Transmission System
Author*Keisuke Osawa, Hiromasa Habuchi (Ibaraki University, Japan), Yusuke Kozawa (Tokyo University of Science, Japan)
Pagepp. 807 - 810
KeywordCode Shift Keying, Visible light communication, Dimming control, Light emitting diode
AbstractIn this paper, a variable N-parallel code shift keying (VN-CSK) with the modified prime sequence code (MPSC) for an illumination light data transmission system is proposed. In the proposed system, LED light is used not only for illuminating rooms but also for data transmission. The bit error rate performances are evaluated by theoretical analysis. Consequently, it is found that the proposed system can actualize the dimming control and data transmission simultaneously. The bit error rate performance for N=1 has better approximately 1.3 [dB] than that for N=4 when BER is 10-6.

W2-4-2 (Time: 10:50 - 11:10)
TitleA Study on ACO-OFDM Systems for Deep-Sea Visible Light Communications
Author*Hiroshi Yamaoka, Masato Saito (University of the Ryukyus, Japan)
Pagepp. 811 - 814
KeywordACO-OFDM, Visible Light Communications, Deep-sea communications
AbstractIn this paper, we consider ACO-OFDM (Asym- metrically Clipped Optical-Orthogonal Frequency Division Multiplexing) based VLC (Visible Light Communication) systems for deep-sea communications. As the deep-sea chan- nel model, we use the model which includes the effects of AWGN (Additive White Gaussian Noise) and attenuation due to water, chlorophyl, and fulvic and humic acid. We evaluate the ACO-OFDM system in terms of BER and average channel capacity.

W2-4-3 (Time: 11:10 - 11:30)
TitleRate-Adaptive LDPC Reconciliation and Estimation in Quantum Key Distribution
AuthorPisit Vanichchanunt (King Mongkut’s University of Technology North Bangkok, Thailand), *Tharathorn Phromsa-ard (Chulalongkorn University, Thailand), Patcharapong Treeviriyanupub (Phranakhon Rajabhat University, Thailand), Paramin Sangwongngam (National Electronics and Computer Technology Center, Thailand), Lunchakorn Wuttisittikulkij (Chulalongkorn University, Thailand)
Pagepp. 815 - 818
KeywordLDPC codes, Rate-adaptive reconciliation, Quantum key distribution, Maximum-likelihood, Reconciliation efficiency
AbstractThe key reconciliation method is proposed as one of the classical parts in Quantum Key Distribution (QKD) protocol. The proposed method aims to correct the transmission error after distribution of quantum key objects over a quantum channel. For error correction, Low-Density Parity-Check (LDPC) codes are adopted as the technique of source coding with side information or the Slepian-Wolf coding. In this work, rate adaptive LDPC reconciliation method based on puncturing and shortening technique with estimated Quantum Bit Error Rate (QBER) from only syndrome is studied. Its objective is to skip the step of quantum bit error rate (QBER) estimation by the traditional key sampling that increases the final secret key length in practice. This method also estimates reconciliation efficiency in advance for determination of an optimal rate that reduces the interactive communication. From numerical results, it can be observed that the performance of our proposed schemes in terms of reconciliation efficiency and secret key throughput is superior to conventional Cascade, and fixed-rate compatible LDPC based protocols. Therefore, gain of these proposed schemes is suitable for the high-speed discrete-variable QKD applications.

W2-4-4 (Time: 11:30 - 11:50)
TitleQuantifying the Benefits of Pulse Shaping for Equalized PAM Transmissions in VLC Systems
AuthorSu Pyae Sone (Asian Institute of Technology, Thailand), *Poompat Saengudomlert (Bangkok University, Thailand)
Pagepp. 819 - 822
Keywordvisible light communications, pulse shaping, pulse amplitude modulation, bit error rate, linear equalization
AbstractIn digital data transmissions, it is known that pulse shaping can help improve the transmission performance by reducing the amount of frequency selective distortion on the transmitted signal. However, to what degree pulse shaping is needed and by how much the transmission performance is improved in the context of visible light communications (VLC) has not explicity been reported. This paper provides numerical results from simulation experiments on data transmissions using pulse amplitude moduation (PAM) over two commonly used channel models for VLC, namely the exponential channel impulse response (CIR) and the ceiling-bounced CIR. For equalization, the linear minimum mean square error (MMSE) criterion is adopted for the sake of a possible practical implementation. For unequalized PAM transmissions, numerical results indicate that pulse shaping using the square-root raised cosine pulse (SRRC) and pulse shaping using the rectangular pulse yield similar bit error rate (BER) performances. However, in the presence of linear MMSE equalization with a fixed filter length, pulse shaping using the SRRC pulse can lead to significantly lower BERs when compared to the baseline case of using the rectangular pulse.


Session W2-6  Audio / Speech Signal Processing (2)
Time: 10:30 - 11:50 Wednesday, July 13, 2016
Location: Room 6
Chair: Naoto Sasaoka (Tottori University, Japan)

W2-6-1 (Time: 10:30 - 10:50)
TitleEvolutionary Stagnation Avoidance for Design of CSD Coefficient FIR Filters Using GA
Author*Yuto Arie, Kenji Suyama (Tokyo Denki University, Japan)
Pagepp. 823 - 826
KeywordFIR filter, GA, combinatorial optimization
AbstractIn this paper, a design method of Canonic Signed Digit (CSD) coefficient Finite Impulse Response (FIR) filters using Genetic Algorithm (GA) is discussed. When applying a general GA, a diversification of search is prevented because the individuals having similar genetic structure increase with increasing search. Therefore, an evolutionary stagnation often occurs. In the proposed method, multiple populations are used to promote the diversification and to avoid the evolutionary stagnation. Several design examples are shown to present the effectiveness of the proposed method.

W2-6-2 (Time: 10:50 - 11:10)
TitleSpeech Enhancement Based on Distribution of Sound Source Localization Results
Author*Naoki Shinohara, Kenji Suyama (Tokyo Denki University, Japan)
Pagepp. 827 - 830
Keywordmicrophone array, sound source enhancement, linear filter
AbstractIn this paper, a target sound enhancement method by using an adaptive microphone array is studied. Then, a supervised signal is required for learning filters. Under a stationary noise environment, it is easy to generate the supervised signal from a power envelope of the received signal. Otherwise, it is difficult to generate the supervised signal under a nonstationary noise environment by the same way. In the proposed method, the supervised signal is generated based on the distribution of sound source localization results. Several experimental results in a real environment are shown to present the effectiveness of the proposed method.

W2-6-3 (Time: 11:10 - 11:30)
TitlePre-inverse Active Noise Control System with Auxiliary Filter
Author*Kazushi Otsuki, Naoto Sasaoka, Ryousuke Minagi, Yoshio Itoh (Graduate School of Engineering Tottori University, Japan)
Pagepp. 831 - 834
Keywordactive noise control, adaptive signal processing
AbstractFiltered-x algorithm is widely used for an active noise control (ANC). However, it has a possibility of making ANC unstable due to the modeling error of secondary path. Therefore, this paper introduces a bias free structure into the estimation of the inverse transfer function of the secondary path. The proposed structure takes advantage of the delay of the primary path.



Monday, July 11, 2016

Session P1  Poster Session I
Time: 13:30 - 15:30 Monday, July 11, 2016
Location: Foyer
Chair: Morikazu Nakamura (University of the Ryukyus, Japan)

P1-1 ([Circuits & Systems] Analog Circuits)
TitleA Mixed-mode Biquadratic Circuit Employing OTAs and Grounded Capacitors
Author*Takao Tsukutani, Masumi Fukuma (Matsue College of Technology, Japan), Yasuaki Sumi (Tottori City Colleage of Medical Nurse, Japan), Noboru Yabuki (Tsuyama College of Technology, Japan)
Pagepp. 835 - 838
KeywordActive circuit, Mixed-mode operation, Biquadratic characteristics, OTAs
AbstractThis paper introduces a mixed-mode biquadratic circuit employing Operational Transconductance Amplifiers(OTAs) and grounded capacitors. The circuit can perform mixed-mode operations by selecting the input and output terminals. And the circuit enables to realize low-pass, band-pass, high-pass, band-stop and all-pass transfer functions suitably choosing the input terminals. Additionally, the circuit parameters ω0 and Q can be tuned orthogonally through adjusting the transconductance gains of the OTAs. An example is given together with simulated results by PSPICE.

P1-2 ([Circuits & Systems] Analog Circuits)
TitleA Self-reference Sensing to Compensate Data Pattern Dependent Leakage Current in the Cross-point ReRAM Array
Author*Hojun Kim, Kee-Won Kwon (Sungkyunkwan University, Republic of Korea)
Pagepp. 839 - 842
KeywordCross-point array, self reference, Data dependent leakage current, Sensing margin, Subtractor
AbstractThis paper presents a self-reference sensing which could compensate data pattern dependent leakage current that acts as a noise in cross-point ReRAM array. There has not been a clear solution for the leakage current issue without using selector devices. Proposed scheme uses analog subtractor to compensate the leakage current. This compensation leads us to get a uniform and enough (over 30mV) sensing margin superior to 2mV of the conventional one in the worst case of data pattern.

P1-3 ([Circuits & Systems] Analog Circuits)
TitleSelf-Terminated Read for Reduced Read Disturbance and Power Consumption in Large Cross-Point ReRAM Arrays
Author*Hyun-Seok Kang, Kee-Won Kwon (College of Information and Communication Engineering, Sungkyunkwan University, Republic of Korea)
Pagepp. 843 - 846
KeywordReRAM, self terminated, read disturb, power, sneak current
AbstractThis paper presents self-terminated read scheme for improving read disturbance and reduction of power consumption in large cross-point ReRAM arrays. The residual read stress prolongs after the decision is made resulting in problems of read disturbance and undesirable power consumption. Elimination of unneccessary read stress cuts the total stress by 80% saving the read power by 40.63% in 1-Mb cross-point ReRAM array. The read disturbance is improved about 10 times compared to conventional scheme.

P1-4 ([Circuits & Systems] Analog Circuits)
TitleA New Linear Low-gm OTA with DC Current Control
AuthorFujihiko Matsumoto, *Shota Matsuo, Takeshi Ohbuchi, Ryuji Yasugi (National Defense Academy of Japan, Japan)
Pagepp. 847 - 850
KeywordOTA, Analog circuis, Low frequency, Active filters, Linear circuits
AbstractThis paper proposes a new linear low-gm OTA. The advantages of the proposed OTA are the simple structure, and easiness of design. In addition, it should be noticed that input signal attenuator with precise coefficient is not necessary. The optimum value of the linearization parameter K is obtained theoretically from maximally flat approximation without relying on numerical analysis. Simulation results show that the linearity of the proposed OTA is about the same as the conventional OTA.

P1-5 ([Circuits & Systems] Analog Circuits)
TitleA Design of a Low Noise and Low Power Low Drop Out Regulator with Fast Settling Time Technique
Author*Jung-Yeon Kim, Seong-Jin Oh, Dong-Soo Lee, Kang-Yoon Lee (College of information and Communication Engineering, Sungkyunkwan University, Republic of Korea)
Pagepp. 851 - 852
KeywordLow noise, Low Drop out Regulator, Fast Settling Time
AbstractThis Paper presents a low noise and low power Low Drop Out regulator (LDO) with fast settling time technique. Using Low Pass Filter (LPF) at the output of Band Gap Reference (BGR) and using output shunt capacitor at the output of LDO are the most efficient ways of suppressing output noise. Its settling time, however, is extremely slow because of large resistance and capacitance. By adopting the proposed fast settling time technique, settling time becomes 6 times faster compared to the normal situation while maintaining outstanding output noise characteristics. This circuit is designed in CMOS 55 nm process and the output noises are 300 nV⁄√Hz, 22 nV⁄√Hz, and 0.21 nV⁄√Hz at 10 kHz, 100 kHz, and 1 MHz, respectively. Settling time of LDO is 150 us and the output of the proposed LDO is 850 mV. The current consumption including BGR and LDO is 82 uA when the supply voltage is 1.2 V

P1-6 ([Computers] Computer Systems & Applications)
TitleA Scheme of Motion Recognition For Wearable Device Gesture
Author*Inhye Park, Sang-Yub Lee, Jae-Jin Ko (Korea Electronics Technology Institute, Republic of Korea)
Pagepp. 853 - 856
KeywordWearable device gesture, Air gesture recognition, MEMS sensors
AbstractThis paper presents a scheme of the motion recognition for wearable device gesture. To recognize motion of user’s arm, we use 6-axis sensors, which consists of a 3-axis accelerometer, a 3-axis gyroscope. When user moves as predefined gesture with wearable devices, the several sensor data forms correlation relation between themselves over the certain level. So, we implement cross-correlation function between generated sensor data. And we filter these sensor data as a certain level to classify specific patterns. We show overall motion recognition architecture at the first. Then our a methodology including the generated sensor data and data processing are explained. Finally, we experiment our method to test platforms divided as wearable device and AVN platform and show the results.

P1-7 ([Circuits & Systems] Other)
TitleBiomimetic Robot Hand Control By Using Surface Electromyography
AuthorStephen Ryan Angsanto, *Jonghun Kwon, Jungho Park, Gwanwoo Kim, Wansu Lim (Kumoh National Institute of Technology, Republic of Korea)
Pagepp. 857 - 860
KeywordElectromyography, Neural Network, Pattern Recognition
AbstractUndesirable events may happen to people at unexpected times. About 15% of the world’s population lives with some form of disability. Most prosthetic devices have limited number of gestures and cannot fully replicate movements of the body. The system utilizes non-invasive surface electromyography (sEMG) for extracting signals from the forearm and uses neural network for pattern recognition. The EMG signals were recorded and calibrated for one participant only. The datasets were sampled to create the input matrix, which are loaded to MATLAB for training, validation, and testing. As shown after successive trials, fatigue or muscle weakness is a significant factor in creating neural networks for pattern recognition. It was verified that the system could successfully extract, classify and output 10 individual finger gestures and 4 manual grasps with a classification accuracy of 93.6%. Statistical analysis was used to assess the classification accuracy based on the results and the original training data with 99% level of confidence.

P1-8 ([Communications] Communication Signal Processing)
TitlePerformance Evaluation of Clusters in Factor Graph for Graph-Based Detection
Author*Thanomsak Sopon (Rajamangala University of Technology Isan, Thailand)
Pagepp. 861 - 864
KeywordGraph-Based Detector, BPMR System, 2-D Interference Channel, Inter-Track Interference, Media Noise
AbstractTwo-dimensional (2-D) interference channel with inter-symbol interference (ISI) and inter-track interference (ITI) can degrade the performance of the read channel in bit patterned media recording (BPMR) system. In this work, we propose the method to improved graph-based (GB) detector. The factor graph applies to clusters in the cluster 2×3 channel matrix on 2-D interference channel to the GB detector. The simulation results show the performance of the proposed GB detector method with the conventional GB detector and multi-SOVA (64s) detector achieve the gains of about 0.8 and 1.0 dB at bit error rate equal 10-4, in the BPMR system with media noise at areal density 2 Tbits/in2.

P1-9 ([Circuits & Systems] Other)
TitleDesign and Preliminary Implementation of an IoT-Based System for Ocean Observation Buoys
Author*Si Moon Kim, Woon Hyun Lee, Hyuk Jin Kwon, Jeongchang Kim (Korea Maritime and Ocean University, Republic of Korea)
Pagepp. 865 - 867
KeywordBuoy, Ocean, Gateway, IoT, CAN
AbstractIn this paper, we propose an IoT-based system for ocean observation buoys. The proposed system consists of various sensor modules, a gateway, and a remote monitoring site. In order to integrate sensor modules with various communication interfaces, we propose a CAN-based a sensor data packet and a protocol for the gateway. The proposed scheme supports the registration and management of sensor modules so as to make the buoy system easier to manage various sensor modules.

P1-10 ([Computers] Computer Systems & Applications)
TitleA Case for PRAM-based NVDIMM
Author*Kyungyeon Yang, Jae W. Lee (Sungkyunkwan University, Republic of Korea)
Pagepp. 869 - 870
KeywordNVDIMM, PRAM, energy efficiency
AbstractNVDIMM has drawn much attention as a solution to back up the main memory at power outage. However, today’s NVDIMM typically uses NAND flash devices, which are not byte-addressable and have poor write performance. This paper proposes PRAM-based NVDIMM (P-NVDIMM), which replaces NAND flash devices with PRAM devices. Since PRAM is byte-addressable with low read latency, we place read-intensive pages into PRAM, instead of DRAM, to save refresh power. Furthermore, the write speed of PRAM is much faster than that of NAND flash to significantly reduce the energy consumption for backup. During normal operations P-NVDIMM saves 29% of energy consumption compared to the flash-based NVDIMM, while incurring only 3.9% of IPC degradation. Besides, P-NVDIMM consumes only 1.22% of energy to back up 8GB DRAM.

P1-11 ([Computers] Computer Vision)
TitleOptical Character Recognition Performance Comparison of Convolution Neural Network and Tesseract
Author*Daegun Ko, Suhan Song, Kimin Kang, Seongwook Han (Samsung Electronics, Republic of Korea), Juneho Yi (Sungkyunkwan University, Republic of Korea)
Pagepp. 871 - 874
KeywordOptical Character Recognition, Tesseract, Deep Learning
AbstractAn OCR is text recognition which is designed to extract ASCII code from an image. Our main idea is to compare two methods in OCR: one is Deep Learning based training system; another is Tesseract based pattern recognition. In this paper, we used Caffe OCR character sets to measure recognition rate and process time between Deep Learning and Tesseract. As a result, we recommend a Deep Learning method due to its performance on recognition rate. However, if process time is on priority, we recommend a Tesseract for its speed

P1-12 ([Computers] Computer Vision)
Title3D Alignment of Face in a Single Image
Author*Eungyeol Song, Kyungjae Lee, Sangyoun Lee (Yonsei University, Republic of Korea)
Pagepp. 875 - 877
Keywordalignment, face landmark, registration, 2d to 3d estimation, 3d alignment
AbstractThe objective of this thesis paper is to suggest a method to find a face and align 3 dimensional model landmark through 67 landmarks that describe characteristics of a face from a single image. Furthermore, using Z-value from the deformable base, a 2 dimensional image can be converted into a 3 dimensional image. The suggested algorithm is the AAM algorithm, which can detect faces precisely, and the Procruest Analysis, which, among other 3 dimensional model lanmarkds, aligns with the most precision. Using the aligned point cloud, ratation, translation, and scale matrix are calculated. A 2D face is converted to a 3D fact using this coefficient.

P1-13 ([Computers] Computer Vision)
TitlePhotometric Stereo with Relit Images from a Single Frontal Face Image
Author*Junhyeop Lee, Kyungjae Lee, Sanyoon Lee (Yonsei University, Republic of Korea)
Pagepp. 879 - 882
KeywordFace image, Photometric stereo, Relighting, 3D Reconstruction, Maximum-likelihood estimation
AbstractConverting 2D images to 3D models using photometric stereo is a long-standing problem. Works on photometric stereo method have been reported that it requires multiple images at least three images upto thousands images with varying poses and illumination direction to generate a depth map. We focus on this issue and propose a solution for photometric stereo with relit face images from a single face frontal image. First, with a relighting method, we can augment face images based on reference images with varying illumination directions. Using the generated virtual faces, relit face images, we can estimate surface normals using the Lambertian model and depth map using surface normals from photometric stereo method.

P1-14 ([Computers] Computer Vision)
TitleLow Light Image Enhancement using Estimated Shade Map
Author*Chang Woo Song, Woo Jin Jeong, Young Shik Moon (Hanyang University, Republic of Korea)
Pagepp. 883 - 886
Keywordcontrast enhancement
AbstractWe propose a method for single image contrast enhancement that operate on image which involving shade area affected by low light condition. First, we estimate shade map in image. Then, for restoration shade area, we apply it to the proposed shading image model. As the result, we obtain an unshaded image. We also provide a comparison of our proposed image enhancement method with state-of-the-art methods by Measure of Enhancement (EME). As a result, we conclude that the best performance achieved.

P1-15 ([Computers] Computer Vision)
TitleRoad Surface Estimation using a Piecewise Linear Function in v-Disparity
Author*Kyungjae Lee, Jaesung Kim, Sangyoun Lee (Yonsei University, Republic of Korea)
Pagepp. 887 - 889
Keywordroad estimation, v-disparity, depth image
AbstractIn this paper we proposed an algorithm to estimate a road surface for a vehicle. Our method utilized a picewise linear function in v-disparity space directly to reduce time consumption without transformation of real-world coordinates. For robust algorithm, we implement a series of image processing procedures. u-disparity is used for filtering obstacles which disturb candidates extraction step of a road surface. v-disparity is used for extraction of a road surface candidates and estimation of a road surface. We separate regions from v-disparity and implement a line fitting algorithm independently. Experimental results show that our method estimates a road surface model successfully.

P1-16 ([Computers] Image Processing)
TitleVisual Salience based Exposure Fusion Algorithm for Ghost Removal in High-dynamic-range Imaging
AuthorZijie Wang, *Qin Liu, Tongwei Ren, Bin Luo (Software Institute,Nanjing University, China), Takeshi Ikenaga (Graduate School of Information, Production and Systems, Waseda University, Japan)
Pagepp. 891 - 894
KeywordHigh Dynamic Range Imaging, Visual Salience, Exposure Fusion, Moving Objects Detection, Ghost Removal
AbstractHigh-dynamic-range Imaging(HDRI) techniques are proposed to present a similar range of luminance as the true image we capture through our visual system. To extend the limited dynamic range of camera sensors, HDRI synthesis techniques merge multi-exposures images. One main problem in HDR image synthesis is removing ghost artifacts caused by moving objects. The proposed algorithm computes visual saliency to constrain the distance calculation for moving areas detection and modifies the exposure fusion method with moving objects masks and saliency maps.

P1-17 ([Computers] Image Processing)
TitlePerformance Analysis of Compressive Image Sensing for Various Spatial Resolutions based on Sparsity Measure
Author*Moonhee Kim, Younghyeon Park, Byeungwoo Jeon (Sungkyunkwan University, Republic of Korea)
Pagepp. 895 - 898
KeywordCompressive sensing, Block-based compressive sensing, Spatial resolution, Sparsity measure, Gini index
AbstractCompressive sensing (CS) can lead sampling with a lower rate than that suggested by the Nyquist/Shannon’s theorem for the sparse signals. The block-based CS (BCS) requires less memory size for the sensing matrix than the frame-based one. Image characteristic can vary according to various spatial resolution of image capturing device, so it is expected to affect the quality of the reconstructed images from compressesively smapled data. We compare objective qualities of reconstructed images of BCS for various image spatial resolutions using Gini index as sparsity measure. The simulation result shows that reconstructed image has better objective quality by increasing spatial resolution.

P1-18 ([Computers] Image Processing)
TitleA Study on Tracking and Trajectory Prediction of High-Speed and Long-Range Flight Target using Infrared Stereo Cameras
Author*Jaeha Lee, Jun Ho Oh (Department of Computer and Software, Hanyang University, Republic of Korea), Sang Hwa Lee (Institute of New Media and Communications, Seoul National University, Republic of Korea), Kayoung Park (Agency for Defense Development, Republic of Korea), Jong-Il Park (Department of Computer and Software, Hanyang University, Republic of Korea)
Pagepp. 899 - 902
Keywordinfrared camera, stereo system, trajectory prediction, regression analysis, genetic algorithm
AbstractIn this study, it suggests the method to track and predict the trajectory of long range target flying at high speed. Generally, it uses the laser or radar in order to gain the three-dimensional position information of target. However, in case of laser, it is very difficult to take aim at target of high speed, and radar is relatively correct, but very expensive. In this study, hereupon, it proposes the way to utilize the heat of flying target as tracking information, using infrared stereo cameras. The suggested method is calculated the three-dimensional position information of long range target, by building the stereo environment with long base line, and it finds the exact position of target in thermal imagery, by finding the leading part of heat source. Also, it predicts next trajectory of target using regression analysis and genetic algorithm.

P1-19 ([Computers] Image Processing)
TitleAn Analysis on Projection Suitability Using Projector-Camera System
AuthorJoongseok Song, *Changseob Kim, Jong-Il Park (Hanyang University, Republic of Korea)
Pagepp. 903 - 906
KeywordProjector, Projector-camera, Plane Analysis
AbstractIn this paper, we propose an analysis on projection suitability using projector-camera system. Firstly, projector projects a horizontal-lines image onto the wall, and camera captures the projection image. Then, we obtain an edge image on the acquired image using Canny edge detector. Through the edge image, proposed system can calculate the total score on the wall using three complexities that are texture, color, and geometric complexity. The geometric complexity is calculated by signs of each slope of the edge in each local area, which is applied Hough-transform.

P1-20 ([Computers] Image Processing)
TitleSuperpixel-Based Approach for View Synthesis Robust to Noise
Author*Joo-Hyeon Kim, Tae-Youn Kim (Hanyang University, Republic of Korea), Hong-Chang Shin (Electronics and Telecommunications Research Institute, Republic of Korea), Jong-Il Park (Hanyang University, Republic of Korea)
Pagepp. 907 - 910
KeywordView Synthesis, Super Pixel, DIBR
AbstractDepth image based rendering (DIBR) technique allows to create a virtual view image by using images and disparity maps captured from different viewpoints. In general, the quality of the synthesized image highly depends on the accuracy of the disparity maps; therefore, the synthesized image has poor quality if noise exists in some region of the disparity map. For solving this problem, we propose a robust method to synthesize a high quality image of virtual view despite of disparity noise. The reference images are segmented into superpixels, and we use the disparity value which is not regarded as noise among disparity values in each superpixel for DIBR in order to suppress the effect of noise. The experimental results shows the effectiveness of the proposed method.

P1-21 ([Communications] Multimedia Communications)
TitleIntroduction of Mobile Edge Computing System Architecture for Media Service on Highway
Author*Sang-Yub Lee, In Hye Park, Jae Kyu Lee (Korea Electronics Technology Institute, Republic of Korea), Seung Yeon Kim (Korea University, Republic of Korea)
Pagepp. 911 - 914
KeywordBus, VOD service, Mobile Edge Computing, Cloud service, Fog computing
AbstractCloud services to smart devices face latency and data traffic issues. Mobile Edge Computing System (MECS) are positioned between cloud and mobile platforms. Their high speed data transmission to users and rapid hand over service from base stations enables real time and location based service. This paper expands this system to the bus on the highway and moving cars along with the bus as special case of MECS, and relates it to the method of media data streaming service with connection continuously. For the realistic vehicular environment, selection algorithm and scheduling method for car requiring streaming media service are proposed and system performance considering vehicular environment is shown

P1-22 ([Communications] Mobile & Wireless Communications)
TitleOrthogonal Frequency and Code Division Multiplexing Using Modified Pseudo-Ternary M-Sequences Pair
Author*Yuki Hosokawa, Hiromasa Habuchi (Ibaraki University, Japan)
Pagepp. 915 - 918
KeywordOFCDM, Code Shift Keying, M-sequence, bi-orthogonal code, peak-to-average power ratio
AbstractIn this paper, Orthogonal Frequency and Code Division Multiplexing(OFCDM) system using Modified Pseudo-Ternary M-sequences(MPTM) pair as the spreading codes is proposed. MPTM is made by mixing {0,1}-valued M-sequences and bi-orthogonal codes. The proposed system achieves the PAPR reduction more than the conventional system without the deterioration of bit error rate(BER) performance. BER of the proposed system is identical to that of the conventional system with the {+1,-1}-valued orthogonal M-sequences.


Session P2  Poster Session II
Time: 15:50 - 17:50 Monday, July 11, 2016
Location: Foyer
Chairs: Thanomsak Sopon (Rajamangala University of Technology Isan, Thailand), Lin Min Min Myint (Shinawatra University, Thailand)

P2-1 ([Communications] Ubiquitous Networks)
TitleProposal of Cloud-Based Women’s Healthcare System Using Ubiquitous Network
Author*Hiroyuki Okazaki, Hiroki Matsumoto (Maebashi Institute of Technology, Japan), Naoyuki Masada (NAKAYO,INC, Japan)
Pagepp. 919 - 922
KeywordUbiquitous network, Cloud service, Tablet telephone, Women’s Healthcare
AbstractIn this paper, we describe such a mechanism, evaluate its effectiveness as the Women's Healthcare by opinion score tests, and the necessity of building a cloud-based Women's Healthcare system which is the integration of both systems and uses ubiquitous network.

P2-2 ([Circuits & Systems] Analog Circuits)
TitleAn Ultra-Low Power and High Resolution Digitally Controlled Oscillator with Gain Estimation Technique for Direct Modulation
Author*Seong-Jin Oh, Hamed Abbasizadeh (College of information and Communication Engineering, Sungkyunkwan University, Republic of Korea), Sang-Sun Yoo (Department of Electrical Engineering, KAIST, Republic of Korea), Kang-Yoon Lee (College of information and Communication Engineering, Sungkyunkwan University, Republic of Korea)
Pagepp. 923 - 924
KeywordAll-Digital Phase Locked Loop, Digitally Controlled Oscillator, Gain Estimation, Direct Modulation
AbstractThis paper presents an ultra-low power and high resolution Digitally Controlled Oscillator (DCO) with gain estimation technique for direct modulation. Three capacitor banks and Delta Sigma Modulator (DSM) are adopted for ultra-high resolution and for the low current consumption, 0.7 V of supply voltage is adopted. Gain estimation technique is adopted for direct modulation. The proposed DCO is designed and fabricated using 55 nm CMOS process and its oscillation Frequency is 4.8 GHz. 1-bit resolution of MSB, LSB, and Fine cap banks are 81.92 MHz, 5.12 MHz, 153.6 kHz, respectively, the nominal phase noise at 1 MHz offset is -116.28 dBc/Hz, and the current consumption is 850 uA when 0.7 V VDD is supplied.

P2-3 ([Circuits & Systems] Computer Aided Design)
TitleMILP-based Scheduling for Clock Latency Minimization in High-level Synthesis
Author*Keisuke Inoue (Kanazawa Technical College, Japan), Mineo Kaneko (JAIST, Japan)
Pagepp. 925 - 928
Keywordhigh-level synthesis, clock latency, MILP
AbstractThe hardware cost of the realization of a clock skew schedule tends to increase due to additional delay elements. Therefore, there is a demand to reduce clock latencies at early design stage. This paper discusses a novel design problem in high-level synthesis, to minimize clock latencies with scheduling, since scheduling highly affects the required clock latencies. Experimental results show that the proposed approach can reduce, the maximum clock latency, and the sum of clock latencies, over the conventional design.

P2-4 ([Circuits & Systems] Linear / Nonlinear Systems)
TitleA PID Controller Design for Dynamic Thermal Management Considering Performance of Mobile Processor
Author*Han Su Kim, Hyun Hak Cho, Chang Min Eun, Ok Hyun Jeong (Sogang University, Republic of Korea)
Pagepp. 929 - 932
KeywordDynamic Thermal Management, PID Controller, Mobile Devices, Thermal Problem, Operating Frequency
AbstractRecently, operating frequency of mobile processor gets higher and it concludes to thermal problems. Therefore, a thermal management system is required to guarantee the reliable performance. This paper proposes Dynamic Thermal Management with PID controller. As the controller highly adopted thermal characteristic of chipset, it can optimally control the temperature. As the result of benchmark tests, the number of data exceeding set point was reduced for 26.01%. Consequently, new DTM considering thermal characteristic relieved thermal problem of the chipset.

P2-5 ([Computers] Image Processing)
TitleVisual Perspective based Exposure Compensation for 360° Virtual Reality Videos
Author*Jae-Il Jung, Yongjin Kim, Jinhyoung Ahn, Sungrae Cho, Minchul Kim, Yongkuk Kim (Samsung Electronics, Republic of Korea)
Pagepp. 933 - 936
Keywordvirtual reality, exposure compensation, 360° video, visual perspective, gain compensation
AbstractWe propose an exposure compensation algorithm for 360° virtual reality video, which considers visual perspective. Conventional algorithms force images to have the same and fixed exposure, but our algorithm finds dynamic exposures according to viewer’s gazing point. We measure the Euler angles between a current gazing point and centers of stitched images on a virtual sphere, and optimize exposure compensation parameters based on the angles. Experimental results show that our algorithm successfully compensates exposure differences while keeping the original brightness.

P2-6 ([Computers] Computer Systems & Applications)
TitleExtension of DIA in the Microprogram Control Unit
AuthorYoung-Jin Oh, *Gi-Yong Song (Chungbuk National University, Republic of Korea)
Pagepp. 937 - 940
Keywordmicroprogram control unit, extension of a DIA, idle cycle, microopreation missing, side effect
AbstractThe CMCU(compositional microprogram control unit) introducing counter into next-state logic and DIA(dedicated input area) that splits control memory into two parts are reviewed, and then an extension of the concept of a DIA is proposed in order to complement the operation of a DIA. With the extension, some limitations such as occurrence of an idle cycle and microoperation missing or duplication in some cases are removed.

P2-7 ([Computers] Computer Systems & Applications)
TitleA Space-Efficient Footprint DRAM Cache
Author*Jongwon Kim, Yongjun Lee, Hakbeom Jang, Jae W. Lee (Sungkyunkwan University, Republic of Korea)
Pagepp. 941 - 942
KeywordDie-stacked DRAM, Memory footprint
AbstractRecently, 3D die-stacked DRAM technologies have been adopted by many processor vendors as a solution to the memory wall problem. To effectively utilize this large in-package memory, there are proposals to use it as a cache. Page-based caches have drawn much attention to reduce the cost of tags by increasing the granularity of caching. However, page-based caches waste cache capacity by over-fetching blocks that are not actually used during the page's lifetime in the DRAM cache. In this paper, we introduce a novel footprint caching technique, which greatly improves the space efficiency of a page-based cache. The key idea is to overlay two sparse pages with many invalide blocks into a single physical page, thus maximizing page utilization. Our cache design improves the IPC by 17.9% and reduces cache miss by 5% over a state-of-the-art footprint cache.

P2-8 ([Computers] Computer Systems & Applications)
TitleAn Effect of Product Lifespan on Markets Using Multi-Agent Similation
Author*Shun Onodera (Graduate School of Science and Technology, Niigata University, Japan), Nao Ito (Toyama National College of Technology, Japan), Yoshinobu Maeda (Graduate School of Science and Technology, Niigata University, Japan)
Pagepp. 943 - 944
KeywordConsumption, Lifespan, Similation, Yasutomi Model
AbstractIn this study, we investigated the effect of the lifespan of products on the consumption of them in the artificial economic market. We constructed a multi-agent simulation (MAS) model proposed by Yasutomi in 1994, and considered both the amount of consumption and the agent utility in the two cases: 1) only short-life products are circulated, and 2) only long-life products are circulated on the market.

P2-9 ([Computers] Computer Systems & Applications)
TitleEffect of Frequency-Dependent Behavior on Bullying in School Using Multi-Agent Simulation Model
Author*Yu Nishida, Yoshinobu Maeda (Graduate School of Science and Technology, Niigata University, Japan)
Pagepp. 945 - 946
KeywordMAS, Frequency-Dependent Behavior, Bullying, Artificail school class model, Simulation
AbstractThe paper addresses simulation results on school bullying and the relevant interactions between classmates. We have proposed the multi-agent simulation (MAS) model composed of many agents mimicking student's behaviors. The bullying is defined as the emergence of isolated agents who have never the same values (formalized attributes of hobbies and interests) of any other agents. In particular, we clarified the effect of sequential frequency-dependent behaviors on the emergence of the bullying (or the victim) using the MAS model. As a result, it was necessary for reducing the number of victims that at least 20% students should stop the frequency-dependent behavior.

P2-10 ([Computers] Computer Systems & Applications)
TitleWeb-Accessible Maritime Vessel Equipment Smart Monitoring and Maintenance Platform Using MySQL Database and JSP
AuthorStephen Ryan Angsanto, *Gyuhyeon Gim, Sung-Phil Heo, Donguk Kwon, Wansu Lim (Kumoh National Institute of Technology, Republic of Korea)
Pagepp. 947 - 949
KeywordMySQL, Database, Smart Monitoring
AbstractThis paper presents a web-accessible database adapted for monitoring and maintaining equipment on-board most ships to engineers for easy monitoring and quick data retrieval. A module retrieves relevant and important data then transmits them wirelessly to a mobile device which sends them to an off-site database for more detailed analysis later. Designed mainly with MySQL 5.7 and JSP which offers better security in data access, low cost, compatibility, and accessibility with different operating systems and computer architectures than MS Access or FileMaker.

P2-11 ([Computers] Computer Vision)
TitleEstimated Position Based on the Camera by a Fingerprint Method for Optimizing 3D Effects
Author*Sung-muk Kang, Jinhoon Park, Yun-su Choi, Jin-won Park, Jun-dong Cho (Sungkyunkwan University, Republic of Korea)
Pagepp. 951 - 954
Keyword3D viewing, Haar-like, HSV, Labeling
AbstractWe present a method for users to effectively watch 3D. There is disparity in the 3D image. The disparity is varied based on the user’s location, the method that we have proposed is possible to change the user’s viewing environment. As a method to know the user’s location, first, it must detect the user’s face in haar-like. Place the area to find detect the user’s face to the region of interest. The region of interest distinguishes the color of the glasses in the HSV color model. Then, we obtain the numbers of pixels glasses. In comparison with the data that has been previously saved, you can keep track of the user’s location.

P2-12 ([Computers] Computer Vision)
TitleFace Frontalization Method using Depth Information
Author*Xingjie Zhang, Joongseok Song, Jong-Il Park (Hanyang University, Republic of Korea)
Pagepp. 955 - 958
Keywordface frontalization, face alighment, face recognition, face identification
AbstractRecently, the face recognition system has achieved excellent results. However, the performance of face recognition is affected by various factors such as expressions, occlusions and pose variations. And the pose variations is the most important one of these. In this paper, we propose a face frontalization technique which used face landmarks and depth information. This method could be applied in the preprocessing stage of the face recognition system and transform the non-fronted face to the front face. In experiments, we verified that a set of faces with various pose could be transformed to the front face by using proposed method.

P2-13 ([Computers] Computer Vision)
TitleRobust Occlusion Recognition Based on Ring Projection Transforms
Author*Yun-su Choi, Jin-hoon Park, Ghulam Hussain, Jong-hak Kim, Jin-won Park, Jun-dong Cho (Sungkyunkwan University, Republic of Korea)
Pagepp. 959 - 962
KeywordRecognition, Ring Projection Transform, Occlusion, Template matching, angle
AbstractIn this paper, we propose a method for locating occluded object and measuring its angle using Ring Projection Transform (RPT) template matching. The current template matching method does not accurately recognize an object that is occluded in a scene image because correlations between the reference template and scene image decrease with occlusion. However, the proposed method divides the template image into sectors to allow partial recognition for locating an occluded object. The angle of the object can also be calculated using the center point of the partially recognized area

P2-14 ([Computers] Computer Vision)
TitleSpectrum based Color Representation on Augmented Reality
Author*Kyudong Sim, Jong-Il Park (Hanyang University, Republic of Korea)
Pagepp. 963 - 966
KeywordAugmented reality, multispectral imaging, color correction
AbstractAs mobile devices have improved, many people have experienced augmented reality. However, an augmented object seems to be floated on real object because of color difference, shadow effect, and so on, and a lot of studies are conducted to overcome the problem. In this paper, we suggest augmented reality based on spectrum in a robust illumination. To display a spectral object on a RGB color display device, power spectrum of illumination and sensitivity of camera are obtained using an expanded marker which contains many colors. Evaluations are conducted by comparing relit color accuracy between conventional color correction and suggested method because the suggested method shows high accuracy and well color adaptation in illumination, it is expected to be used in various illumination condition.

P2-15 ([Computers] Computer Vision)
TitleMap Operation Interface using Thermal Imaging Camera
Author*Kanghoon Lee (Hanyang University, Republic of Korea), Taeyoung Uhm (Korean Institute of Science and Technology, Republic of Korea), Jong-Il Park (Hanyang University, Republic of Korea)
Pagepp. 967 - 970
KeywordThermal Image Camera, Residual Heat, Fingertip Touching, Thermography, Touch Detection
AbstractIn this paper, we propose a thermal infrared image based map operation interface system that can be examined by merging a historical map and a current map. This system consists of the fusion of both maps and finger touch detection.

P2-16 ([Computers] Image Processing)
TitleVideo Analysis Algorithm based on Saliency Region Detection from Selected Key-frames
Author*Min Soo Ko, Hyok Song (Korea Electronics Technology Institute, Republic of Korea)
Pagepp. 971 - 974
KeywordVideo analysis, Saliency map, Scene change detection, Keyframe selection
AbstractIn this paper, more effective algorithms to extract content-aware information from the video is proposed.In the past few years, the amount of the internet video has grown rapidly, and it has become a major market. Reliable extraction of metadata from the video is one major step toward efficient video management. Extraction of reliable and useful metadata from a video involves video understanding and content analysis, and it is always a tall order in terms of both algorithms and system design. In the proposed algorithm, we can extract the four kinds of information such as image category, camera motion, saliency region and keyframe from a user’s video.

P2-17 ([Computers] Motion Analysis)
TitleRelative Position Feature based Dense Trajectories with Density Adapted Noise Reduction for Tennis Player Action Recognition
Author*Zihan Ma, Shuyi Huang, Masaaki Honda, Takeshi Ikenaga (Waseda University, Japan)
Pagepp. 975 - 978
Keywordaction recognition, tennis sports analysis, Dense Trajectories, noise reduction, relative position
AbstractTennis player action recognition plays an important role in tennis sports analysis. The outdoor environment noise and high similarity features of arm motion at different torso sides are two crucial problems in tennis player action recognition. This paper proposes a density adapted noise reduction and a relative position feature to recognize player actions. The density adapted noise reduction adaptively removes sparse noise far away from the player region, according to the variable density of features in each frame. The relative position feature makes features at different sides of the player torso distinguishable to differ backhand receive action from serve and forehand receive actions. Experiments executed on sequences of a practical tennis game achieve accuracy higher than 93.8%, and AUC value larger than 0.94 for all action categories.

P2-18 ([Computers] Multimedia Service & Technology)
TitleCloud-based Real-Time Streaming Service for UHD Contents
Author*Yo Seph Choi, Seongsin Jeon, Woo Chool Park (Korea Electronics Technology Institute, Republic of Korea)
Pagepp. 979 - 982
KeywordStreaming, HEVC, UHD
AbstractDigital broadcasting technology has been to equip the environment that can service UHD contents. At this time video streaming service is a very popular technology in which these services strive continuously transmit video content from the server to the client. Video delivery technology to deliver the best quality to the user automatically has become a very important challenge at this point. By using the Hypertext Transfer Protocol (HTTP) which is the de facto protocol of the Internet service at this time, a new streaming method has been carried out research and development.

P2-19 ([Computers] Security)
TitleA Design of Hybrid Honeypot System for Malicious Code Security
Author*Moongoo Lee (Kimpo University, Republic of Korea)
Pagepp. 983 - 986
Keywordmalicious, Honeypot, internal, preemptive, attack
AbstractIn order to protect information asset from various malicious code, Honeypot system is implemented. Honeypot system is designed to have attack eliciting purpose in order to not accept internal system attack, or it is designed to purpose of malicious code information collection. Therefore, this study offers hybrid Honeypot that offers internal information system security function and information collection for malicious security. This provides a security preemptive response effect about malicious code but also information security analyze for recent hacking techniques and attacking tool as well.

P2-20 ([Computers] Security)
TitleA Design of Smart Home Security System
Author*Hyoung-Ro Lee, Jun-Young Jang, Chi-Ho Lin (Semyung University, Republic of Korea)
Pagepp. 987 - 990
KeywordHome Security, Ultrasonic sensor, Arduino, Motion sensor
AbstractIn this paper, we propose a Smart home security system. Proposed smart home security system consists of arduino which is inexpensice main processor, ultrasonic sensor and human body detection sensor to detect whether someone breaks into home. Data from ultrasonic and human body detection sensor are transmitted to web server via ethernet shield connected to ardudion. Web server checks whether someone breaks into home by using stored data from ultrasonic sensor and human body detection sensor. Snapshot is photographed via webcam connected, user can monitor in web or smart device environment. When examining effciency of proposed smart home security system, it was found that proposed system is easier to be made than existing home security system, is cost effective by using arduino and is effcient and convenient and stable as it enables a user to handle an error in person and it uses reliable data.

P2-21 ([Computers] Computer Systems & Applications)
TitleA David-Star Magic Square Algorithm for Efficient LED Control
Author*Kyung-Min Lee, Su-Jeoung Yun, Chi-Ho Lin (Semyung University, Republic of Korea)
Pagepp. 991 - 994
KeywordLED, lighting, algorithm, a magic square, Illuminance distribution
AbstractIn this paper, we propose a David-Star Magic Square Algorithm for Efiicient LED Control. Existing Lighting system has reduced power consumption and Heat. And it increased the life of the LED module. However light fixtures, there is a limit in accordance with the form and balance of illuminance distribution. Proposed a David-Star Magic square algorithm for efficient LED control is efficience of power consumption, Heat, module life and illuminance distribytion. This paper, the proposed algoritme is arranged twelve sub-module of David-Star frame in 16x16 LED DotMatrix. And the proposed method create a pattern through a specific time interval. Comparison of the conventional lighting system and method using a magic square increased the efficiency of the control



Tuesday, July 12, 2016

Session P3  Poster Session III
Time: 13:30 - 15:30 Tuesday, July 12, 2016
Location: Foyer
Chair: Cheon Won Choi (Dankook University, Republic of Korea)

P3-1 ([Circuits & Systems] Semiconductor Devices and Technology)
TitleAnalysis of I/0 Clamp, Power Clamp ESD Protection Circuit
Author*Byung-Seok Lee, Jun-Geol Park, Chung-Kwang Lee, Kyoung-il Do, Yong-Seo Koo (Dankook University, Republic of Korea)
Pagepp. 995 - 998
KeywordESD, SCR, GGNMOS, Trigger voltage, holding voltage
AbstractThis paper presents a ESD protection device for I/O clamp and power clamp. The proposed ESD protection devices has a fast turn on time, low trigger voltage, and high holding voltage characteristics than conventional ESD protection device. The proposed device was analyzed to figure out electrical characteristics and tolerance robustness in term of individual design parameters (D1, D2, D3). They are investigated by using the Synopsys TCAD simulator. The results show that the STNMOS (Substrate Triggered NMOS) device has lower trigger voltage 4.8V compared to the conventional GGNMOS. In addition the proposed a novel SCR (Silicon Controlled Rectifier) - based ESD (Electrostatic Discharge) protection device for power clamp. The proposed device has a higher holding voltage characteristic than conventional SCR. These characteristics enable to have latch-up immunity under normal operating conditions as well as superior full chip ESD protection. As a result of simulation, holding voltage increased with different design parameters. The holding voltage of the proposed device changes from 3.3V to 7.9V.

P3-2 ([Circuits & Systems] Modern Control)
TitleDisturbance Response and Stability Analysis of Wireless Tele-Control System for MIMO Plant
Author*Faramarz Alsharif, Shiro Tamaki (University of the Ryukyus, Japan), Katsumi Yamashita (Osaka Prefecture University, Japan), Mohammad Reza Alsharif, Bruno Senzio-Savino, Mahdi Khosravy (University of the Ryukyus, Japan), Heung Gyoon Ryu (Chungbuk National University, Republic of Korea)
Pagepp. 999 - 1002
KeywordControl System, Disturbance response, Multipath channel, Adaptive filter, Equalizer
AbstractIn this research we aim to analyze the disturbance response of MIMO plant and wireless communication system. In order to realize a distant control system wireless communication can be applied to control system that contain channels. The effects of channels are reduced by equalizer. Moreover we consider there are disturbances. The plant is set to be a drone that aims to control the attitude. In conclusion we discussed about the disturbances response against drone that is in stationary status.

P3-3 ([Circuits & Systems] VLSI Design)
TitleDesign of Multi-Bin-Per-Cycle HEVC CABAC Encoder with Multiple Bypass Bin Engines
Author*Doohwan Kim, Sohyun Kim, Jeonhak Moon, Seongsoo Lee (School of Electronic Engineering, Soongsil University, Republic of Korea)
Pagepp. 1003 - 1006
KeywordHEVC, CABAC, multi-bin, bypass
AbstractCABAC (Context Adaptive Binary Arithmetic Coding) is high-compression-ratio entropy encoding in HEVC. However, it is difficult to implement fast CABAC encoder, since each bin should be serially processed due to data dependencies of bins. There are two types of bins, i.e. regular bin and bypass bin. Processing time of bypass bin is shorter than that of regular bin, so it is possible to cascade bypass bin engines for multi-bin-per-cycle processing. Number of bypass bin engines is carefully chosen based on thorough analysis. The implemented CABAC encoder can process up to 1.92 bins per cycle.

P3-4 ([Circuits & Systems] VLSI Design)
TitleThe Design of the IP with High Efficiency for IoT using Data Separation Prewitt Edge Detection
Author*Chang-hee Park, Jin-sung Jung, Sang-bok Cho (Department of Electronic Engineering, University of Ulsan, Republic of Korea)
Pagepp. 1007 - 1009
KeywordEdge, detection, prewitt
AbstractIn this paper, in order to implement the IP of edge detection algorithm for surveillance in image using the camera in CCTV or vehicle black box, we designed pre-processing step that is the edge extraction algorithm. First, after input image converts into the input signals of R, G, and B, three inputs are combined, and converted to gray scale. Also, these data is divided in half and apply to each Prewitt edge[1] based on X, Y axis through 3 line buffer, the output image is determined edge image through a combination of two result. A proposed algorithms was implemented using matlab program. This was converted and comfirmed to verilog through xilinx ISE

P3-5 ([Circuits & Systems] VLSI Design)
TitleDesign of Wireless Communication Modem for Internet-of-Things Applications
Author*Hyunsu Hwang, Yongchul Jung, Yunho Jung (Korea Aerospace University, Republic of Korea)
Pagepp. 1011 - 1014
KeywordDSSS, IEEE802.15.4, Internet-of-Things, Wireless Modem, ZigBee
AbstractIn this paper, wireless communication modem for internet-of-things (IoT) applications is proposed which can support various data rates from 31.25Kbps to 2Mbps, and its implementation results are presented. Repetition coding for 32-chip direct-sequence spread spectrum (DSSS) symbol is applied for low rates under 250Kbps to extend the coverage. Convolution coding, puncturing, and interleaving for non-DSSS symbol are performed for high rates from 500Kbps to 2Mbps for multi-media services. Simulation results show that the coverage increases at the rate of 51.8-77.3% for various environments compared with IEEE 802.15.4 ZigBee. The modem chip was implemented in 180nm CMOS process and total gate counts are 260K with the size of 5.8mm2.

P3-6 ([Circuits & Systems] Other)
TitleTunable Aperture controlled by Liquid Crystal Optical Switch
Author*Yumee Kim, Kukjin Chun (Seoul National University, Republic of Korea)
Pagepp. 1015 - 1017
KeywordTunable aperture, Liquid crystal, Ranging and Imaging, Range sensor, optical switch
AbstractWe propose the tunable aperture using liquid crystal (LC). The device size is 9 × 9 × 1.8mm3 and the aperture diameters are 1.35mm and 0.63mm for F3.0 and F6.3 respectively. This device has not only electrical benefits from 3V low driving voltage and 16ms fast response time for the proper mobile application, but also mechanical benefits from absence of movable parts for high reliability and low spherical aberration of circular pattern.

P3-7 ([Computers] Computer Systems & Applications)
TitleSpark vs. Virtualized Spark: A Performance Analysis
Author*Wenjing Jin, Jae W. Lee (Sungkyunkwan University, Suwon, Republic of Korea)
Pagepp. 1019 - 1020
KeywordApache Spark, OpenStack, Virtulization
AbstractApache Spark is an open-source framework for scalable big data processing. OpenStack is a popular virtualization framework that provides Infrastructure as a Service (IaaS) on cloud. Deploying Spark on OpenStack provides many benefits such as on-demand resource scaling, greater availability and flexibility. However, this virtualized Spark is likely to have very different performance characteristics from the native Spark. This paper aims to quantize the cost of virtualization on a Spark cluster. Our experiments demonstrate that (i) the virtualized Spark with four nodes is about 1.58X slower than the native Spark, (ii) all of network, CPU and GC cause this slowdown. Overall, the network waiting time and CPU time contribute the most to the increased execution time, and the GC time has the highest increasing rate.

P3-8 ([Communications] Circuits & Components for Communications)
TitleNon-Multiplier QPSK Modulator Using Simple Trigonometric Function
Author*Lerson Kirasamuthranon, Jeerasuda Koseeyaporn, Pramote Wardkein (King Mongkut's Institute of Technology Ladkrabang, Thailand)
Pagepp. 1021 - 1024
KeywordQPSK, Phase shift keying, Trigonometry
AbstractThe quadrature phase-shift keying (QPSK) is one of the digital modulation techniques using four different phases in order to represent the data bits. Building the QPSK modulating circuit is somewhat complex because of the need of two local frequencies whose phases need to be exactly 90 degree differences. A new technique, which uses just basic Op-Amp circuits in order to create the amplifier circuit and the summing amplifier circuit and also simplifies the QPSK modulation system, is proposed. The designed amplifier circuit is controlled by data bit. Consequently, phase shifting of QPSK signal is regulated by data bit. The data bits are applied so as to control the designed amplifier circuit’s gain which also regulated the QPSK signal’s phases. From this reason, the proposed QPSK modulating circuit is therefore less complicate than the conventional one. Both theoretical and experimental results thus confirm that the propose technique can be realized in real world application.

P3-9 ([Computers] Computer Systems & Applications)
TitleA Cache Design to Optimize Memory Performance in Mobile Proceesor
Author*Su In Kim, Chang Min Eun, Hyun Hak Cho, Ok Hyun Jeong (Sogang University, Republic of Korea)
Pagepp. 1025 - 1028
KeywordCache, Memory Performance, AMAT
AbstractIn recent years, the processor performance has been improved rapidly, but the enhancement speed of memory performance is slow. Due to this gap, the improvement of memory performance is significant, and enhancing the cache performance can assist that. For this reason, we designed the cache that optimizes memory performance in the target processor, Cortex-A57. To evaluate memory and processor performance, Average Memory Access Time (AMAT) and Cycles Per Instruction (CPI) were used, and the simulation was conducted by using SimpleScalar3.0 and benchmarks from SPEC CPU2000. As a result, the performance was enhanced by 34.93% in AMAT and 10.05% in CPI with proposed design which is optimized to target processor. It is contributed to performance improvement of mobile processor as well.

P3-10 ([Computers] Computer Vision)
TitleIris Feature Extraction using Local Thresholding with Variable Block Size
Author*Ryo Ishikawa, Tomohiko Ohtsuka, Hiroyuki Aoki, Yuji Tateizumi (National Institute of Technology, Tokyo College, Japan)
Pagepp. 1029 - 1032
Keywordiris feature extraction, local thresholding, variable block size, iris recognition, biometrics
AbstractThis paper presents a novel approach that uses local thresholding with an optimal block size to extract the features of the space domain from an image of the iris. The block size is selected optimally from two kinds, based on the variance of pixel intensity in the block of interest. The results of several experiments show that a complete separation of the matching intra-class and inter-class score distribution can be achieved. Furthermore, the equal error rate and false rejection rate achieved by using the proposed approach are 0.02% and 0.32%, respectively, thereby improving those obtained with the conventional approach, i.e., 0.89% and 24%, respectively.

P3-11 ([Computers] Image Coding & Analysis)
TitleFour-directional Edge-Effect Method of Measuring in Laser-printers
Author*Daegun Ko, Suhan Song, Yousun Bang, Kimin Kang, Seongwook Han (Samsung Electronics, Republic of Korea)
Pagepp. 1033 - 1035
Keywordprint image quality, white-gapping, edge-effect, optical density fading
AbstractEdge-Effect is printing optical density fading phenomenon which is white-gapping between solid and halftone of black area. In this paper, we propose a new method for measuring Edge-Effect based on psychophysical experiments and data analysis method. Our approach is to extract two components both Edge-Effect’s integral (PA) and white-peak (PB). Edge-Effect is computed by the weighted sum of PA and PB, where two weighting parameters are determined by psychophysical experiments. By conducting psychophysical validation experiments, the strong correlation is obtained between the proposed measuring method and the perceived scales. The correlation coefficient r is 0.95.

P3-12 ([Computers] Image Coding & Analysis)
TitleThe UHD Content Packet Format for Distributed Transcoding
Author*Yo Seph Choi, Seongsin Jeon, Woo Chool Park (Korea Electronics Technology Institute, Republic of Korea)
Pagepp. 1037 - 1040
KeywordUHD, Transcoding
AbstractThe extra high-efficient codec technology that based on the HEVC is essential for streaming the UHD content over the Internet. It requires an extra highefficiency transcoder technology to simultaneously support various resolutions because it expected to have SD, HD, UHD content consumption environments coexist in the Nscreen. Therefore, from the existing low-efficiency compression format (H.264 / AVC, VC - 1, such as ProRes) to the extra high-efficiency multi- resolution images that is convertible to (SD, HD, UHD) compression format from the transcoder described is essential. At domestic and abroad are currently in progress to develop technology for real-time transcoding UHD, however the UHD Ultra- Capacity Due to the characteristics of the situations that require research and development of cloud-based distributed transcoder technology. In this paper, we propose a Packet Analyzer and KMSP packet structure to perform effectively distributed transcoding.

P3-13 ([Computers] Image Coding & Analysis)
TitleOptimization of a Tone-mapping Curve for High Dynamic Range (HDR) Video
Author*Chanyup Chae, Seounghwi Kim, Dongkyu Lee, Jongseok Lee (Kwangwoon University, Republic of Korea), Jungwon Kang (Electronics and Telecommunications Research Institute, Republic of Korea), Seoungjun Oh (Kwangwoon University, Republic of Korea)
Pagepp. 1041 - 1044
KeywordHDR video compression, Tone-mapping, LDR image quality, Adaptive histogram segmentation
AbstractTone-mapping is the important part in the backward compatible structure of a video coding. While tonemapping an HDR (High Dynamic Range) image/video, bitrate and LDR (Low Dynamic Range) image/video quality are mainly considerable issues. Some pioneer approaches deal with the HDR bitrate rather than LDR image/video quality. In this paper, a tone-mapping curve is proposed with an adaptive histogram segmentation method to preserve LDR image quality while minimizing HDR-MSE (Mean Square Error). Then, we adjust the tone-mapping curve to make the LDR image have higher contrast and be more natural. In our experiment, the results show the subjectively better LDR image and objectively higher values of TMQI (Tone-Mapping Quality Index) metric.

P3-14 ([Computers] Image Coding & Analysis)
TitleVisual Fatigue Detection Using Binocular Disparity and PPG in Stereoscopy
Author*Jin-hoon Park, Ghulam Hussain, Yun-su Choi, Jong-hak Kim, Jin-won Park, Jun-dong Cho (Sungkyunkwan University, Republic of Korea)
Pagepp. 1045 - 1048
Keywordvisual fatigue, depth map, stereoscopic, Photo plethysmography
AbstractThe amount of visual fatigue caused by stereoscopic videos and its corresponding physiological changes are analyzed. We analyze visual fatigue using binocular disparity which is the most fundamental cause of such fatigue. The previous researches used a methods of comparing data obtained before and after watching 3D videos; however, the comparative data analysis from such method is hard to find. This research aims to achieve relevant results that can be gathered by determining the VFL (Visual Fatigue Level) of 3D videos and analyzing the changes in body signals while watching the videos.

P3-15 ([Computers] Image Processing)
TitleFire Extinguisher Recognition using Robust Color Detection and Gamma Correction
Author*Hyun Tae Kim, Young Min Jang, Sang Bock Cho (Department of Electronics Engineering, University of Ulsan, Republic of Korea)
Pagepp. 1049 - 1052
Keywordfire extinguisher detection, color detection, gamma correction
AbstractIn this paper, we propose an algorithm to detect the fire extinguisher in real time through the camera lens using the robust color detection for reducing optical noise with color detection method and gamma correction to 2-channel operation. The proposed algorithm has been decreased the calculation time about 64% (conventional method : 4.182ms, proposed method : 1.505ms) than conventional method using prominent object color to change the YCbCr model. Also, the error rate by the noise decreases about 43% because proposed method uses a gamma correction. As a result, it shows the processing speed of 10.8ms per frame and the detection rate of 96% per frame, it can quickly and accurately detect the color information for object tracking.

P3-16 ([Computers] Security)
TitleGaussian Filtering Detection of Digital Image Forensic
AuthorSe Hwan Park (KISTI ReSEAT Program, Republic of Korea), *Kang Hyeon Rhee (Dept. of Electronics Eng., Chosun University, Republic of Korea)
Pagepp. 1053 - 1056
KeywordDigital Image Forensic, Gaussian Filtering Detection
AbstractFor a design of the Gaussian filtering (GF) detection (GFD) in the altered digital images, this paper presents a new feature vector that is formed from the autoregressive (AR) coefficients by AR model of the gradients of the horizontal and vertical lines in an image. In the proposed algorithm, AR coefficients are computed from the gradients of the lines. Subsequently, the defined 20-dim. feature vector is trained in a SVM (Support Vector Machine) for the GFD in the forged images. In the experiment, two kinds test items are the area under curve (AUC), and a minimal average decision error. The performance is excellent both at GF (3×3) vs. median filtering (3×3), and GF (5×5) vs. original and JPEG (90) on the DFD. However, in the measured performances of the AUC by the sensitivity (TP: True Positive rate) and 1-specificity (FP: False Positive rate) is above 0.9. Thus, it is confirmed that the grade evaluation of the proposed algorithm is rated as “Excellent (A).”

P3-17 ([Computers] Other)
TitleA Low-Cost Hybrid SSD Architecture for Read-Intensive Enterprise Workloads
Author*Hyeon Gyu Cho (Sungkyunkwan University, Republic of Korea), Kyungho Shin (Sungkyunkwan University, Samsung Electronics, Republic of Korea), Jaeyoung Jang, Jae W. Lee (Sungkyunkwan University, Republic of Korea)
Pagepp. 1057 - 1058
KeywordPCM, Hybrid SSD, Enterprise workload
AbstractRecntly, flash-based solid-state drives (SSDs) have been widely adopted in all scales of computing platforms due to their low latency and high energy efficiency. Many enterprise workloads are known to be read-intensive, and read latency has great impact on overall system performance. This paper proposes ROSA, a novel PCM-Flash hybrid SSD architecture optimized for read-intensive enterprise workloads.

P3-18 ([Communications] Antenna & Wave Propagation)
TitleDesign of Dual-Band MIMO Antennas with Isolation Enhancement
Author*Ching-Lieh Li, Yu-Feng Shih, Yen-Ting Wu (Tamkang University, Taiwan)
Pagepp. 1059 - 1060
Keyworddual-band antenna, MIMO antenna, isolation, WLAN
AbstractIn this paper, we propose a planar dual-band MIMO antenna with improved isolation for WLAN applications (2.45 GHz and 5~6 GHz). The MIMO antenna consists of two antenna elements that are perpendicularly placed in order to achieve the polarization diversity for the high band, such that the coupling between the antenna elements can be reduced. On the other hand, additional slot cut on the common ground plane is introduced to further enhance the isolation for the low band. Simulated and experimental results show that the proposed techniques can effectively inhibit the coupling.

P3-19 ([Communications] Audio / Speech Signal Processing)
TitleStudy on Detection Scheme of Direction of Ship Whistle in the Frequency Domain
Author*Hyukjin Kwon, Jeongchang Kim (Korea Maritime and Ocean University, Republic of Korea)
Pagepp. 1061 - 1063
KeywordDSP, Whistle signal, Ship, SRS
AbstractIn this paper, we propose a sound reception system (SRS) using a phase difference of whistle signals from four microphones installed on a ship. The phase difference is obtained in the frequency domain. The proposed system detects the direction of the received whistle signal and indicates the direction of other ships. Also, we implement the SRS using a digital signal processor chip and verify the operation of the proposed SRS.

P3-20 ([Communications] MIMO & Space-Time Codes)
TitleA Zero-forcing Tomlinson-Harashima Precoding with Stream Permutation based on the Bit Rate Maximization for Multi-user MIMO Systems
AuthorKazuki Koiso (Hiroshima University, Japan), *Shigenori Kinjo (Japan Coast Guard Academy, Japan), Shuichi Ohno (Hiroshima University, Japan)
Pagepp. 1065 - 1068
KeywordMU-MIMO, SDM, zero forcing, THP, MMSE
AbstractThe purpose of this report is to develop a zero-forcing Tomlinson-Harashima precoding scheme for the multi-user MIMO scenario. The algorithm and methodology of the single-user MIMO case are the base of our proposal. Finally, we execute the feasibility study of the proposed method.

P3-21 ([Communications] Other)
TitleImproved Decoding of Binary Cyclic Codes
Author*Chong-Dao Lee, Yaotsu Chang (I-Shou University, Taiwan)
Pagepp. 1069 - 1072
KeywordCyclic codes, Decoding, Extended Euclidean algorithm, Unknown syndromes
AbstractRecently, the unified unknown syndrome representations have been used to decode binary cyclic codes up to actual minimum distances. Two practical methods to determine unknown syndromes as sparse polynomials in terms of syndromes are proposed by using the syndrome matrices. Experimental results show that the presented methods provide the sparse polynomials for the unknown syndromes and the developed decoders based on them can significantly reduce both delay time and area overhead.

P3-22 ([Communications] Other)
TitleClassification of Attention Deficit/Hyperactivity Disorder (ADHD) by Extracting Non-linear Features of Children's EEG
AuthorAhmadreza Heidarpour, Mousa Shamsi (Sahand University of Technology, Iran), Faramarz Alsharif, *Bruno Senzio-Savino, Mohammad Reza Alsharif (University of the Ryukyus, Japan)
Pagepp. 1073 - 1075
KeywordAttention deficit/hyperactivity disorder (ADHD), Entropy, Empirical mode decomposition (EMD), Wilcoxon test, Support victor machine(SVM)
AbstractADHD is the most frequent disorder in children. According to researchers, it is the most common disorder of childhood and detection of that is very important. In this study, database includes totally 30 subjects, of which 12 are the ADHD and the other 18 are healthy subjects. Using non-linear features, that selected by Wilcoxon test, we address the classification of both groups (healthy and hyperactive). In this paper, the support victor machine (SVM) classifier with 4 kernel function (Polynomial kernel_3, Multilayer perceptron(mlp), Radial basis function(rbf), and quadratic) is used.