Day 1 26/11/2019
Room #1

Registration 09:00 - 09:30

Opening 09:30 - 09:45

Dr. Po Tao Chen & Prof. Imrich Chlamtac & Prof. Li-Fung Chang

Keynote Speech by Prof. I.F. Akyildiz 09:45 - 10:45

Chaired by Prof. Jyh-Cheng Chen

Break 10:45 - 11:00

Keynote by Prof. Yi-Bing Li 11:00 - 12:00

Chaired by Prof. Imrich Chlamtac

Lunch 12:00 - 13:30

Technical Sessions 1 13:30 - 15:00

Meeting Room: East, Session chaired by Prof. Rung-Shiang Cheng
13:30 - 13:45
A Deep Reinforcement Learning Based Intrusion Detection System (DRL-IDS) For Securing Wireless Sensor Networks and Internet of Things

Many modern infrastructures incorporate a number of sen- sors and actuators interconnected via wireless links using Wireless Sensor Network (WSN) and Internet of Things (IoT) technology. With a num- ber of mission-critical infrastructures embracing these technologies, the security of such infrastructures assumes paramount importance. A mo- tivated malicious adversary, if not kept in check by a strong defense, can cause much damage in such settings by taking actions that compromise the availability, integrity, con dentiality of network services as well as the privacy of users. This motivates the development of a strong Intrusion Detection System (IDS). In this paper, we have proposed a new Deep Reinforcement Learning (DRL)-based IDS for WSNs and IoTs that uses the formalism of Markov decision process (MDP) to improve the IDS de- cision performance. To evaluate the performance of our scheme, we com- pare our scheme against the baseline benchmark of standard reinforce- ment learning (RL) and the supervised algorithm of machine learning K-Nearest Neighbors (KNN). Through our a thorough simulation-based performance analysis, we demonstrate that our model DRL-IDS returns superior performance in terms of improved detection rate and enhance- ment the production of accuracy with reduced number of false alarms compared with this current approaches.
Authors: Abderrahim Benslimane (University of Avignon), Hafsa Benaddi (Ibn Tofail University, Faculty of Sciences, MISC Lab, Morocco), Khalil Ibrahimi (Ibn Tofail University, Faculty of Sciences, MISC Lab, Morocco), Junaid Qadir (Information Technology University, Lahore, Pakistan),
Hide Authors & Abstract

Show Authors & Abstract
13:45 - 14:00
BatTalk: Monitoring Asian Parti-colored Bats through the IoT Technology

In the past, we studied the activities of Asian parti-colored (APC) bats through visual observation, which is very labor intensive. This paper develops an IoT platform called BatTalk to continuously monitor the APC bats population, understand its compositional changes, life history, and environmental factors. With BatTalk, the above visual observations can be achieved with reduced man power and minimal interference to the bat activities. The most important task is to use BatTalk to automate the process to understand APC bats’ regular annual cycle of life history and estimate the percentage of the baby bats born and raised there would return to their original habitat in the coming year. Also, we proposed an inexpensive manner to identify the bat habitats and bat movement paths by identifying the bat’s ultrasonic signal strength and GPS position, and then show the information in a map.
Authors: Yun-Wei Lin (College of Artificial Intelligence and Green Energy, National Chiao Tung University, Taiwan), Cheng-Han Chou (Department of Forestry and Natural Resources, National Chiayi University, Taiwan), Yi-Bing Lin (Department of Computer Science, National Chiao Tung University, Taiwan), Wen-Shu Lai (Institute of Applied Arts, National Chiao Tung University, Taiwan),
Hide Authors & Abstract

Show Authors & Abstract
14:00 - 14:15
A Group Signature Scheme for Securing Blockchain-based Mobile Edge Computing

Blockchain-based mobile edge computing (BMEC) is a promising architecture in the fifth-generation (5G) networks. BMEC solves the problem of limited computing resources of devices in the mobile blockchain environment, while ensuring the distributed deployment of computing resources and the traceable of transaction data. However, some consensus-level security threats exist in mobile blockchain environment, i.e., double-spend attacks, long-range attacks, selfish mining. All of these threats can break the integrity of BMEC, allowing the correct block record to be overwritten with a false one. In this paper, we propose a group signature scheme on blocks of blockchain for addressing such issues. Each new block will be regarded as a valid block if it obtains a valid group aggregate signature of the group which the block creator belongs to. The security analysis is also presented to prove that our proposed group signature scheme is effective.
Authors: Shijie Zhang (Sangmyung University, Republic of Korea), Jong-Hyouk Lee (Sangmyung University, Republic of Korea),
Hide Authors & Abstract

Show Authors & Abstract
14:15 - 14:30
VANET Data Offloading using the Mobile Edge Computing (MEC) Paradigm and Architecture

This work proposed a time-based k-hop V2V routing for Mobile Edge Computing (MEC)-based VANET data offloading in the urban vehicular environment. The main idea is to allow a vehicle to enable data offloading from the cellular network to the IEEE 802.11p network when there is a k-hop V2V path connecting the vehicle with its ahead IEEE 802.11p Road Side Unit (RSU). An MEC server, which can collect periodically reported contexts from those vehicles that are inside the administered region, is used to compute whether the k-hop VANET data offloading can be enabled between the vehicle and its ahead RSU or not. The performance analysis shown that the proposed method can increase the data offloading fraction than the traditional data offloading method, which has the data offloading when the vehicle is inside RSU’s signal coverage, in different data offloading scenarios of different vehicular density situations.
Authors: Chung-Ming Huang (Department of Computer Science and Information Engineering, National Cheng Kung University), Shih-Yang Lin (School of Transportation and Vehicle Engineering, Shandong University of Technology, Shan Dong, China), Zhong-You Wu (Dept. of Computer Science and Information Engineering National Cheng Kung University, Tainan, Taiwan),
Hide Authors & Abstract

Show Authors & Abstract
14:45 - 14:45
Towards the Implementation of Movie Recommender System by Using Unsupervised Machine Learning Schemes

Abstract. This study aimed at finding out the similarities within groups of people to build a movie recommending system for users. The recommendation system is very useful for customers to choose the preferred movie with existing features. In this study, the development of a recommendation system is carried out by using some algorithms to get clusterings such as K-Means Algorithm, Birch Algorithm, Mini Batch K-Means Algorithm, Mean shift Algorithm, and Affinity Propagation Algorithm. We proposed a method to optimize K that for each cluster would not rise significantly the variance. We limited to use clustering based on Genre, Tags, and movies ratings. This study would find a better method and way to evaluate clustering algorithm. To verify the quality of the recommendation system, we employed the mean squared error (MSE) and social network analysis (SNA) to explore the relationships between clusters, such as Degree Centrality, Closeness Centrality, and Betweenness Centrality. We also used average similarity, computational time, and clustering performance evaluation as evaluation measures which have been widely used to compare methods performance of recommendation systems. Clustering Performance Evaluation with Silhouette Coefficient, Calinski-Harabaz Index, Davies-Bouldin Index.
Authors: Debby Putri (National Taiwan University of Science and Technology), Jenq Shiou Leu (National Taiwan University of Science and Technology),
Hide Authors & Abstract

Show Authors & Abstract
14:45 - 15:00
Combination of OFDM and CDMA Techniques for a High Bandwidth Optimization and a Great Improve-ment of Signal Quality in OFDM Systems

The use of OFDM modulation has become a priority in recent years as it is an ideal platform for wireless data transmissions. Its implementation can be seen in most of the newer broadband and high bit rate wireless systems, including Wi-Fi, cellular telecommunications and more. This is due to the many benefits this technology provides. These include immunity to selective fading, intersymbol in-terference resistance, intercarrier interference resistance, more efficient spectrum utilization and simpler channel equalization. But with the increasing demand from users, the scarcity of the radio spectrum and the use of OFDM for a large number of recent wireless applications, the optimiza-tion of bandwidth and signal quality must be a major concern. Thus, in this paper, we propose in a multipath Additive White Gaussian Noise en-vironment, a more efficient wireless transmission system that combines OFDM and CDMA techniques. It is a 4-QAM OFDM-CDMA synchronous multiuser sys-tem that uses OVSF codes to differentiate users. It can be applied on the downlink of a wireless cellular system based on a simple OFDM access and even be a system for 5th generation mobiles where OFDM is considered in combination with a multiple access technique.
Authors: Agnès NGOM (Ecole Supérieure polytechnique),
Hide Authors & Abstract

Show Authors & Abstract

Break 15:00 - 15:30

Technical Sessions 3 15:30 - 17:00

Meeting Room: East, Session chaired by Prof. Rung-Shiang Cheng
15:30 - 15:45
Flag-assisted Early Release of RRC Scheme for Power Saving in NB-IoT System

As the 5G standard is about to be completed, the IoT applications will have the unprecedented development. Massive and various IoT devices sense and interact with the environment. Most of those devices are battery-powered and some of them are deployed at inaccessible locations, so how to reduce the power consumption is a critical issue. 3GPP proposed eDRX and two transport optimization mechanisms to help devices reduce power consumption. In this paper, based on the CP CIoT EPS Optimization, a flag-assisted Early Release of RRC scheme is proposed. The message flow is modified to make IoT devices enter RRC IDLE early. The result shows that the flag-assisted Early Release of RRC can help IoT devices save power further.
Authors: Hui-Ling Chang (National Cheng Kung University), Chung-Ying Hsieh (National Cheng Kung University), Meng-Hsun Tsai,
Hide Authors & Abstract

Show Authors & Abstract
15:45 - 16:00
Energy Management for Zones-based Isolated DC Multi-microgrids

In this paper, zones based distributed energy management for isolated multi-microgrids is proposed. Loads are categorized into different zones to form zonal multi-microgrids. Each microgrid has own energy management system which can locally manages the supply and demand of its zonal load and minimize the operational cost. Distributed Network Operator (DNO) act as central controller to facilitate the energy exchange between zonal microgrids and balancing the overall systemwide supply and demand in economic way. Demand Response Program (DRP) is also utilized for peak load shifting within the scheduling horizon. In addition to minimization of operation cost, the utilization of DR will also assure the reliability of supply. In the proposed distributed energy management, each microgrid balances its supply and demand locally and exchange surplus and deficit power with other microgrids through DNO. The performance of proposed scheme is demonstrated through case study simulation of radial multi-microgrid structure.
Authors: Jing Wu (Shanghai Jiao Tong University), Arshad Nawaz (Shanghai Jiao Tong University), Chengnian Long (Shanghai Jiao Tong University), Yibing Lin (National Chiao Tung University),
Hide Authors & Abstract

Show Authors & Abstract
16:00 - 16:15
IoT Insider Attack - Survey

The "Internet of things" (IoT) creating a perfect storm in the smart world. Due to the availability of internet and capabilities of devices, sensors-based technologies becoming popular day by day. In our daily life, IOT now widely used which includes transportation, health, education, security and so on. It now opens the opportunities of overcoming many new challenges. We can’t imagine how gradually it creates impact in our daily lives. Any device with on/off capability if connecting through internet via sensor can be IOT device which includes coffee machine, light, hand watch, headphones, washing machine, mobile phones, car, CCTV camera and so on. Simply we can say connecting things to people via internet and controlling remotely is the great advantages of IOT. Imagine how IOT can make our life easier, based on your set alarm when you wake up, if it can notify your coffee machine to prepare coffee for you that will save your time. Despite those advantages, IOT based system is not free from vulnerabilities. Different types of attack make the system vulnerable and tried to exploit the system and creating obstacle from its growing. Here we will try to explore the possible insider attack and the relevant technologies along with machine learning strategy to overcome those obstacles.
Authors: Morshed Chowdhury (Deakin University - Melbourne), Robin Doss (Deakin University), Biplob Ray (Central Qeensland University-Australia), Sutharshan Rajasegarar (Deakin University- Australia),
Hide Authors & Abstract

Show Authors & Abstract
16:15 - 16:30
HomeTalk: A Smart Home Platform

This paper utilizes an IoT platform called IoTtalk to shape “consideration” into a house to make it a smart home. The developed project is called HomeTalk that serves as a platform to accommodate various smart applications in a house. We describe the following HomeTalk applications. The PlantTalk application takes care of house plants. The FishTalk application provides fish comfortable life in the fish tank at home. The BreathTalk application detects the number of people in a room, which also purify the air. The TheaterTalk application uses home and special appliances to create the effects for a 4D experience theater at home. The FrameTalk application allows a painting frame to interact with people in a house. The GardenTalk application provides smart gardening. Then we show how these applications share the sensors and actuators in the house. In the future, we will integrate them through an award winning project called Orchid House.
Authors: Yi-Bing Lin (College of Computer Science, National Chiao Tung University), Yun-Wei Lin (College of Artificial Intelligence and Green Energy, National Chiao Tung University, Taiwan), Sheng-Kai Tseng (Graduate Institute of Architecture, National Chiao Tung University, Taiwan), Jyun-Kai Liao (Department of Computer Science, National Chiao Tung University, Taiwan), Ta-Hsien Hsu (Department of Computer Science, National Chiao Tung University, Taiwan),
Hide Authors & Abstract

Show Authors & Abstract
16:30 - 16:45
Deep Learning Based Pest Identification on Mobile

Crops, vegetables, fruit trees, flowers and other cash crops, are often harmed by a variety of harmful organisms, plant pathogens, pests, weeds and pest rats, etc. Plant diseases and insect pests often occur, which are one of the main factors which causes the damage of leaves and crop failure. Therefore, in order to stop the pest, it is extremely important to correctly identify the pests of plants and their characteristics. In this paper, an effective and scalable image recognition algorithm is proposed for disease detection. Meanwhile, MobileNets is employed for developing our method on mobile devices. Finally, the effectiveness of our method is demonstrated by experiments using a dataset of apple disease taken in northwest China. In the experiments, transfer learning is used to train a deep convolutional neural network for identifying two types of pest damage, apple rusts and apple Alternaria leaf spot. Our results show that the MobileNets model offer a fast, affordable, and easily deployable strategy for digital plant disease detection.
Authors: Chongke Bi (College of Intelligence and Computing, Tianjin University), Yulin Duan (Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences), Dandan Li (Beijing Agristrong Science & Technology Development Co.,Ltd),
Hide Authors & Abstract

Show Authors & Abstract
16:45 - 17:00
Update Using multi-channel transmission technology to construct an IoT mechanism for homecare service

Owing to the issue of silver tsunami, the number of widow and widower arises day by day. Patient’s families could not accompany elders due to the job. Thus, the elders’ self-care ability becomes an ordeal in daily lives. Even if the advanced medical tech allows elders to have a perfect medical service or to enhance his/her self-care ability, the medical institution is still facing the heavy-burden predicament because of the short of medical manpower and the restriction of medical resource while most caring services are concentrated in hospital or institution, and fail to decentralize them to various home environments. Whereas the target of “Aging in place” sets in the Long-term Care strategy, and to relieve the pressure of Chinese-type treatment and the hardship of long journey. Nevertheless, the popularizing performance is confined due to lack of self-care environment and professional integrating care platform for elders. We intend to use the module of medical internet to conduct clinical field simulation and deployment, through multiple transit technique and fog-computing environment to produce an appropriate aging-care module.It is livable for patients’ health, and can save the unnecessarily medical resource and the manpower cost expenditure. Such a module can be extended to broaden the range of medical service, introduction of smart high-tech, create a livable environment for elders’ healthcare and life.
Authors: Lun-Ping Hung (National Taipei University of Nursing and Health Sciences), Shih-Chieh Li (National Taipei University of Nursing and Health Sciences), Kuan-Yang Chen (National Taipei University of Nursing and Health Sciences), Chien-Liang Chen (Overseas Chinese University),
Hide Authors & Abstract

Show Authors & Abstract

Gala Dinner 18:30 - 21:00

Room #2

Technical Sessions 2 13:30 - 15:00

Meeting Room: West, Session chaired by Prof. Yu-Liang Liu
13:30 - 13:45
Network Protocols and Connectivity for Internet of Things

The aim of the paper is to compare different network protocols which are available for Internet of Things (IoT) systems in different industries and also to define the best practices for using these protocols in different IoT applications. IoT is a huge ecosystem of connected smart devices and objects which gathers enormous amount of data that needs to be captured, processed and communicated to and from the cloud system. The network protocols are compared based on the different IoT systems architecture and connections including smart object to object (O2O), smart object to gateway, gateway to data centers and between data centers. Furthermore, they are grouped based upon different network ranges and network topologies. Selecting the best protocol for IoT application is centered upon the proposed three-dimensional network design model which equates each of the communication protocols against the three axes of the model which are the battery life, device duty cycle and device to gateway range.
Authors: Manan Bawa (false), Dagmar Caganova (Slovak University of Technology in Bratislava Faculty of Materials Science and Technology in Trnava),
Hide Authors & Abstract

Show Authors & Abstract
13:45 - 14:00
An Edge Computing Architecture for Object Detection

Edge computing services are contingent on several constraints. There is a requirement needed to provide a proper function, such as low latency, low energy consumption, and high performance. Object detection analysis involves high power resources, it is because of the need to process the images or videos. In this paper, the architecture of edge computing for object recognition is proposed, and the performance of the edge node is examined. The resources performance comparison on Raspberry Pi and Neural Compute Stick are inspected. This study combined the Neural Compute Stick (NCS) to enhance the ability of image processing on Raspberry Pi. Through the aid of NCS, the Raspberry Pi’s frames per second (FPS) is increased by four times when the object detection program is executed, and the energy consumption of the Raspberry Pi is also recorded.
Authors: Endah Kristiani (Tunghai University), Po-Cheng Ko (Tunghai University), Chao-Tung Yang (Tunghai University), Chin-Yin Huang (Tunghai University),
Hide Authors & Abstract

Show Authors & Abstract
14:00 - 14:15
The Implementation of an Edge Computing Architecture with LoRaWAN for Air Quality Monitoring Applications

Cloud computing enables a user to access and analysis the data at any time, anywhere, and any devices with internet access. However, the need for faster and more reliable cannot adequately be handled by cloud computing. In this case, cloud computing requires edge computing capability to reduce data transport, saving network bandwidth costs, and avoiding data storage proliferation. Therefore, by combining cloud computing and edge computing along with low power wide area networks (LoRaWAN) can provide excellent services. In this paper, a campus air quality using edge computing monitoring system and integrated Arduino and LoRaWAN air quality sensor was proposed. In the Edge layer, we use Raspberry Pi3 with Kubernetes and Docker installation as a hardware device. MySQL database is implemented as a data storage. In the Cloud layer, we use OpenStack clusters for infrastructure virtualization. Ceph is utilized as a distributed storage system for data backup processing. The air quality monitoring data collected by the LoRaWAN sensor is visualized using a web page to monitor and analyze the real-time air pollution data. The LSTM deep learning algorithm is applied to predict air pollution. The air quality data obtained from the open government data and LoRaWAN sensors.
Authors: Endah Kristiani (Tunghai University), Chao-Tung Yang (Tunghai University), Chin-Yin Huang (Tunghai University), Po-Cheng Ko (Tunghai University),
Hide Authors & Abstract

Show Authors & Abstract
14:15 - 14:30
Relay Selection exploiting Genetic Algorithms for Multi-Hop Device-to-Device Communication

Device-to-device (D2D) communication allows a direct transmission between two devices. In this way, cellular user equipment’s are not always obliged to route the data conventionally through a cellular base station. This paper focuses on multi-hop D2D communication, where D2D relays are exploited to delivery of data from a source to a destination. We propose a novel algorithm that finds the most suitable path between the D2D source and destination so that the capacity of multi-hop communication is maximized. The appropriate route is found via Genetic Algorithm (GA) with an ordered crossover. The simulation results show that the proposed algorithm improves the capacity of multi-hop D2D communication from a source to a destination compared to an existing relay selection algorithm by 20–61%. We also show that the proposed solution converges fast enough to be beneficial even in realistic mobile networks.
Authors: Toha Ardi Nugraha (Czech Technical University in Prague), Zdenek Becvar (Czech Technical University in Prague), Pavel Mach (Czech Technical University in Prague),
Hide Authors & Abstract

Show Authors & Abstract
14:30 - 14:45
Design and Implementation of Automatic Following Technology for Mobile Devices

Along with the flourishing development of Internet of Things, vehicles which assist move goods are developed as well, for example, automatic guided vehicle applied in manufacturing plants. Vehicles vary with pattern of goods to be moved. If it moves along magnetic tapes, it would lose its flexibility in moving directions. To make vehicles more dynamic and convenient, this study designs and implements automatic following technology of vehicles. Through relative position be-tween vehicles and objects to be followed positioned by satellite and laser radar installed on vehicles which can detect relative distance, vehicles are able to automatically follow objects to be followed.
Authors: Ming-Fong Tsai (Department of Electronic Engineering, National United University), Chih-Sheng Li (Professional Master’s Program of Information and Electrical Engineering, Feng Chia University), Chi-Feng Chen (Industrial Ph.D. Program of Internet of Things, Feng Chia University), Lien-Wu Chen (Department of Information Engineering and Computer Science, Feng Chia University), Chow-Yen-Desmond Sim (Department of Electrical Engineering, Feng Chia University),
Hide Authors & Abstract

Show Authors & Abstract
14:45 - 15:00
A Plug-in Framework for Efficient Multicast Using SDN

The great variety of modern networked applications, e.g., online computer games, cloud host backups, videoconferencing, etc. brings significant differences in their usage scenarios. Therefore, they impose very different QoS (Quality of Service) requirements on network communication. In particular, traditional multicast implementations cannot react adequately to the potentially very dynamic application requirements at run time. In this paper, we suggest a novel Plug-in Multicast Framework (PiMF) placed on top of an existing multicast framework. PiMF can modify the topology of the multicast tree during the application’s run time, thus providing QoS guarantees for multicast communication. We design our plug-in framework using the emerging SDN (Software-Defined Networking) technology, and we especially address the challenge of non-interfering behavior of PiMF with respect to the underlying multicast implementation. We evaluate the correctness and performance of our plug-in framework in detailed simulation experiments.
Authors: Yu Zhang (University of Muenster), Sergei Gorlatch (University of Muenster), Tim Humernbrum (University of Muenster),
Hide Authors & Abstract

Show Authors & Abstract

Break 15:00 - 15:30

Technical Sessions 4 15:30 - 17:00

Meeting Room: West, Session chaired by Prof. Chien-Liang Chen
15:30 - 15:45
A Coin Recognition System Towards Unmanned Stores Using Convolutional Neural Networks

In unmanned stores, automated checkout is an integral part of the process, and the checkout is usually completed by expensive identification machines. Some un-manned stores lacking banknotes and coins only provide credit cards, EasyCard, or QR code payment methods, sometimes that cause the difficulty of payment when they check out. This research is aimed at the coin recognition for images. It processes the images using OpenCV, and substitutes into the trained convolu-tional neural network (CNN) for identification. The result of the research shows that the accuracy of the model identification is 94%, and it can be used to identify more than one coin.
Authors: Anthony Y. Chang (Overseas Chinese University), Chi Han Chen (Overseas Chinese University), Bo Han Chen (Overseas Chinese University),
Hide Authors & Abstract

Show Authors & Abstract
15:45 - 16:00
Efficient Deployment Based on Congestion Control Delivery In Multi-Hop Wireless Sensor Network

Multi-hop Wireless Sensor networks are designed for various real-time appli-cations, and it should be adequately designed to avoid delivery congestion be-tween sensor nodes. In order to provide an energy efficient transmission, our proposed work introduces a novel optimized method based on the congestion control algorithm. In the proposed approach based congestion control algo-rithm was based on the cluster-based routing, since the energy consumption was effectively reduced throughout the network. Since it has to improve net-work lifetime for a significant simulation period. The delivery control process also reduces the end to end delay. Initially, cluster the nodes with the k means algorithm. After that focusing on the delivery control using Kalman Filter strategy, and this is suitable for high packet delivery prediction. Finally, pack-ets are sending with maximum throughput using optimization-based routing. The simulation is performed on the windows simulation platform, and finally, the performances are evaluated regarding an average end-to-end delay, aver-age throughput, packet delivery ratio, energy efficiency, energy consumption, and reliability.
Authors: Chien-Liang Chen (Department of Innovative Living Design, Overseas Chinese University.), Ding-Jung Chiang (Department of Digital Multimedia Design, Taipei City University of Science and Technology.),
Hide Authors & Abstract

Show Authors & Abstract
16:00 - 16:15
DTMFTalk: A DTMF-based Realization of IoT Remote Control for Smart Elderly Care

Smart elderly care becomes a popular technology to remotely assist the aged at home in recent years because of the population ageing. This paper demonstrates the DTMF-based realization of IoT remote control for telecommunication operators to achieve the smart elderly-care service. By utilizing IoTtalk, we implement a system, DTMFTalk, which supports IoT remote control via conventional circuit-switched DTMF signaling during a phone call conversation. DTMFTalk can monitor the Call State of the smart phone, capture DTMF keys from the in progress call, and send the key values to the IoTtalk server. Afterwards, the smart elderly-care devices can gain the related IoT instructions from the IoTtalk server. Through the delay measurement experiment for DTMFTalk, we observe that DTMFTalk can constantly and accurately recognize the DTMF keys as long as the user holds the desired DTMF keys with enough period.
Authors: Shih-Chun Yuan (National Tsing Hua University), Shun-Ren Yang (National Tsing Hua University), I-Fen Yang (National Tsing Hua University), Yi-Chun Lin (National Tsing Hua University), Yi-Bing Lin (National Chiao Tung University),
Hide Authors & Abstract

Show Authors & Abstract
16:15 - 16:30
Discover the Optimal IoT Packets Routing Path of Software-Defined Network via Artificial Bee Colony Algorithm

The wireless sensor network is the core of the Internet of Things. However, wireless sensors have some limitations and challenges, such as limited power and computing power, data storage, and network bandwidth, especially power requirements. How to find a way to program more flexible and faster according to the state of each sensing node in the network becomes an important issue. The software-defined network separates the control functions from the hardware de-vices, such as switches or routers, so that these hardware devices only have the data forwarding function, and the control software dynamically controls the flow of the network and data packets according to the network state and application requirements. In order to provide flexibility and adaptability, software-defined net-works require a dynamic approach to solving and optimizing routing planning problems. This study will use the artificial bee colony algorithm to monitor the state of the sensor nodes in the software-defined network through the controller and take the best decision dynamically. Artificial bee colony algorithms are used to optimize wireless sensor networks and improve sensor node energy usage and data routing issues. The contribution of this research is to dynamically find the optimal routing path for the sensing nodes through the artificial bee colony algorithm, and improve the overall practicability and reliability of the wireless sensor network.
Authors: Chih-Kun Ke (National Taichung University of Science and Technology), Mei-Yu Wu (National Taichung University of Science and Technology), Wang-Hsin Hsu (National Taichung University of Science and Technology), Chia-Yu Chen (National Taichung University of Science and Technology),
Hide Authors & Abstract

Show Authors & Abstract
16:30 - 16:45
Understanding Mobile User Intent Using Attentive Sequence-to-Sequence RNNs

Smartphones have become an indispensable part of our lives. Understanding user behaviors based on smartphone usage data is therefore critical to many applications. In this paper, we propose to address a novel task called Intention2Text which attempts to capture user intents based on smartphone usage log. The goal of Intention2Text is to learn a deep learning model taking mobile context logs as input and generate sentences as output for describing mobile user intentions. So far, we have developed an attentive sequence-to-sequence recurrent neural network for the Intention2Text task as a fundamental model. Also, various model encoding/decoding strategies are introduced and considered. The experiments based on a real community question dataset are conducted to verify the effectiveness of the proposed framework.
Authors: Che-Hsuan Yu (Chunghwa Telecom Research Institute, Taiwan), Hung-Yuan Chen (ITRI, Taiwn), Fang-Yie Leu (Tung-Hai University), YAOCHUNG Fan (National Chung Hsing University),
Hide Authors & Abstract

Show Authors & Abstract
16:45 - 17:00
IOT Applications to the Land Acquisition Evaluation Mechanism

The construction of the land expropriation assessment mechanism is of great importance in land expropriation. This study conducts a questionnaire survey of scholars and experts in various assessment levels stipulated by the Taiwan Land Acquisition Regulations and ranks the order by weights in the hierarchical analysis method of information application. By expressing the preferred order of factors and facets, the results of the study can provide a reference for subsequent development.
Authors: HUAN-HSIANG LUO (CHUNG HUA UNIVERSITY), Yee-Chaur Lee (Department of Landscape Architecture, College of Architecture and Design, Chung Hua University),
Hide Authors & Abstract

Show Authors & Abstract
Day 2 27/11/2019
Room #1

Keynote Speech by Prof. Mohammed Atiquzzaman 09:30 - 10:30

Chaired by Prof. Abderrahim Benslimane

Break 10:30 - 11:00

Keynote Speech by Assoc. Prof. Dagmar Caganova 11:00 - 12:00

Chaired by Prof. Der-Jiunn Deng

Lunch 12:00 - 13:30

Technical Sessions 5 13:30 - 15:00

Meeting Room: East, Session chaired by Prof. Chun-Hsien Sung
13:30 - 13:45
A Study of Big Data Analytics on Service Quality Evaluation of Online Bookstore

Big Data Analytics can be applied to asses service quality for e-commerce industry achieving customer relationship improvement and reflecting the service quality of transaction. Mosr researches are focus on expert survey or questionnaire-base research to measure the weight and relationship of critical criteria. The goal of this study is to explore and demonstrate the utility of big data analytics by using it to study core online bookstore service quality varia-bles that have been extensively studied in past decades. A text analytical ap-proach to a mining large quantity of consumer reviews extracted from Amazon to deconstruct bookstore customer’s experience and examine its association with satisfaction ratings. This paper proposd a initial study about framework that integration of big data analytic and SERVQUAL model to measure the im-portance and relationship of service quality criteria. Because the service itself has intangible characteristics that are difficult to measure, consumers are more ambiguous in assessing service quality and subjective judgment. Furthrt re-search will apply Amazon data set to evaluate the criteria.
Authors: Ye Xiu Wen (Yulin Normal University), Tsai Jich-Yan (Yulin Normal University), Wang Chien-Hua (Yuan Ze, University),
Hide Authors & Abstract

Show Authors & Abstract
13:45 - 14:00
Cross-border E-commerce Intellectual Property Protection in China-U.S. Trade Friction

Along with the fast development of Internet technology, cross-border E-commerce is also exhibiting an extremely strong growth trend. Among the many problems derived from this new type of cross-border trade, the intellectual property issue is even more complex, and there are many difficulties emerging in governance and rights protection. These difficulties lie in the fact that the conflict between E-commerce and regional intellectual property protection also comes from the lag of technological innovation and legal regulation, with conflicts arising from inadequate international coordination and other realistic causes. The snags in intellectual property protection are also due to the conceptual factor of the conflict between efficiency and fairness. Thus, in order to effectively solve these difficulties, this paper holds that the government, cross-border E-commerce enterprises, intellectual property holders, civil collective forces, and governance bodies at home and abroad should conduct diversified collaboration with one another. In addition, relevant data and an expert talent database should be established, a framework for cross-border E-commerce intellectual property governance and protection must be jointly built, and greater efforts can be made for the sound development of cross-border E-commerce.
Authors: Zhou Ping Ying (Baise University), Ye Xiuwen (Yulin Normal University),
Hide Authors & Abstract

Show Authors & Abstract
14:00 - 14:15
Wavelength assignment for conducting exchanged folded hypercube communication pattern on optical bus

The (s+t+1)-dimensional exchanged folded hypercube, denoted by EFH(s,t), is a brand-new interconnection network proposed by Qi et al. Besides, the bus topology is the simplest topology in optical networks, which can be modeled by a linear array graph, denoted by LA_n. In this paper, the Routing and Wavelength Assignment (RWA) problem for conducting EFH(s,t) communication pattern on LA_n is investigate, where n = s + t + 1. To address this problem, an embedding scheme as well as a wavelength assignment algorithm are proposed. The author also shown that the number of wavelengths required by the wavelength assignment algorithm is 2^(s+t)+2^(s+t-2)+floor(2^t/3).
Authors: Yu-Liang Liu,
Hide Authors & Abstract

Show Authors & Abstract
14:15 - 14:30
Application of Blockchain in Smart Contracts Technology

A smart contract is an agreement whose execution is automated. This automatic execution is often effected through a computer running code that has translated legal prose into an executable program. What makes these legal agreements innovative is that their execution is made automatic through the use of computers. This Article examines smart contracts from a legal perspective. Specifically, this Article explains smart contracts’ operation and place in existing contract law. Blockchains are databases of information that are created by a network with no central authority. For instance, instead of a public recordation system that exists on paper files stored in city hall, a blockchain system would keep a decentralized ledger on the computers of every node running the software. A smart contract does not rely on the state for enforcement, but is a way for contracting parties to ensure performance. For legal purposes, I will further differentiate between strong and weak smart contracts. Strong smart contracts have prohibitive costs of revocation and modification, while weak smart contracts do not. This means that if a court is able to alter a contract after it has been executed with relative ease, then it will be defined as a weak smart contract.
Authors: Chunhsien Sung (Department of Financial and Economic Law, Overseas Chinese University, TW), Chen-Pin Chang (General Education Center, Hsuan Chuang University, Hsinchu, Taiwan),
Hide Authors & Abstract

Show Authors & Abstract
14:30 - 14:45
Power Prediction via Module Temperature for Solar Modules Under Soiling Conditions

The ability to predict the output power of remote solar modules is key to successful wide-scale adoption of solar power. However, solar power is a di-rect product of its environment and can vary vastly from one location to an-other. Predicting generated power for a specific facility requires monitoring the output of the solar modules in the context of ambient variables such as temperature, humidity, solar irradiance, air dust, and wind. This is especially challenging in areas where soiling is a significant environmental variable. Soiling particles such as sand and dust can shade segments of the solar mod-ule, thus effectively reducing the amount of solar irradiance absorbed and, consequently, the power produced. Measuring soiling particles requires ex-pensive equipment that can increase the cost of running the facility and therefore lower the total output. However, dust can also serve as a cooling layer that can reduce the temperature of the solar module and to a certain ex-tent, reduce overheating. This observation can be used to correlate the amount of dust accumulated on the surface of the panel with its temperature. In this work, the module temperature and power output of a clean module and a dusty module are observed using an Internet of Things monitoring sys-tem. The data is used to train various machine learning and deep learning al-gorithms to eventually predict the output of a soiled module over time using only its temperature and a reference clean module.
Authors: Salsabeel Shapsough (American University of Sharjah), Rached Dhaouadi (American University of Sharjah), Imran Zualkernan (American University of Sharjah), Mohannad Takrouri (American University of Sharjah),
Hide Authors & Abstract

Show Authors & Abstract

Break 15:00 - 15:30

Technical Sessions 7 15:30 - 17:00

Meeting Room: East, Session chaired by Prof. Chun-Hsien Sung
15:30 - 15:45
A Research on Blockchain-Based Central Bank Digital Currency

In recent years, the emergence of blockchain-based and privately issued digital currencies has raised a lot of concerns, including: infringement of privacy rights, the danger of such currencies being used as money laundering tools or in a way that harms consumer protection and financial stability. However, central banks have already started their research on the Central Bank Digital Currency (CBDC). To study the subject of CBDC development in China, this paper first presents a detailed introduction of the concept of private digital currency and the issues that come with it. Secondly, this paper advocates the method of establishing an easy-to-regulate CBDC system based on the two chains scheme of blockchain and making sure the complete transaction information and those used for verification are stored and accessed separately, therefore realizing a balance between protection of user privacy and facilitating regulation. At the same time, the consortium blockchain should be anchored in the public chain to ensure data credibility. Furthermore, although China has begun its CBDC development, it has yet to develop adequate laws and regulations. To this end, in addition to presenting a summary of China's CBDC system, this paper also explains the rights and obligations of the central bank, the commercial banks and the public with regards to the currency, in the hope that such contents can be of some help to the revision of relevant laws in the future.
Authors: Cheng-yong Liu (Beijing Institute of Technology, Zhuhai, Guangdong), Chih-Chun Hou (Yulin Normal University, Guangxi),
Hide Authors & Abstract

Show Authors & Abstract
15:45 - 16:00
Intention of online consumers to buy behavior affected by extreme weather-take the mainland Chinese Haining market as an example

The purpose of this study was to investigate whether extreme weather affects online consumer buying behavior. Extreme weather including heavy rain, snow, hot weather and other conditions, whether consumers are unwilling to go shopping, through online shopping directly to buy. Through big data analysis, this paper analyzes whether consumers' online shopping will affect consumers' purchase intention because of extreme weather. We investigated the market in China through a questionnaire survey, and the number of samples we investigated was N = 471. Through the rule correlation analysis method, we found the relationship between the relevant factors to prove it. We found that extreme weather can affect consumers to go shopping to further promote consumers' online shopping.
Authors: Li-Wei Lin (Zhejiang University of Finance & Economics Dongfang College), Shih-Yung Wei (Yulin Normal University), Su-Rong Yan (Zhejiang University of Finance & Economics Dongfang College), Chih-Chun Hou (Yulin Normal University), Shia-Yang Tzeng (Shantou University),
Hide Authors & Abstract

Show Authors & Abstract
16:00 - 16:15
Identifying Smartphone Users Based on Gait Activity Using LSTM Neural Networks

With a wide range applications of smartphones in people’s daily life, the need of secured access control becomes more and more urgent since people tend to store their private and important information, such as personal identifiers and bank account details on their smartphones. In this work, we present a novel framework for user identification technique based on human gait related activities via mobile sensing data. We propose using LSTM Neural Network which is suitable to automate feature extraction from raw sensor inputs for human activity recognition. Furthermore, we investigate the impact of activity dependent identification which produces better accuracy than activity independent identification. The results show that the proposed model is promising compared to other traditional machine learning classifiers and outperforming some of the previous reported results.
Authors: zhang min (jimei university),
Hide Authors & Abstract

Show Authors & Abstract
16:15 - 16:30
Mining network security holes based on data flow analysis technology

With the popularity of mobile terminals and the sharp increase in network data traffic, the problem of security loopholes has become increasingly prominent. The traditional vulnerability detection methods can no longer meet people's demand for detection efficiency. In order to meet the needs of network security and vulnerability detection in the era of big data under the power industrial control system, it is urgent to design a vulnerability mining technology combining big data analysis technology. This paper describes the current situation of big data and vulnerability mining technology and introduces relevant big data security technology and algorithm in detail. The decision tree algorithm is selected as the basic algorithm of big data security technology. Through the simulation test of the decision tree algorithm, the decision tree algorithm is preliminarily tested. The missing alarm rate and false alarm rate of decision tree algorithm are simulated experimentally. Through controlling variables, we obtained the results of three groups of experiments. It proves that our algorithm can effectively detect IP scan, Port scan and other attacks.
Authors: Yang Li (State Grid Xinjiang Electric Power Research Institute), Xiaohua Liu (State Grid Xinjiang Electric Power CO.,LTD.), Lixin Zhang (State Grid Xinjiang Electric Power CO.,LTD.), Winbin Guo (State Grid Xinjiang Electric Power CO.,LTD.), Qian Guo (Global Energy Internet Research Institute Ltd.),
Hide Authors & Abstract

Show Authors & Abstract
16:30 - 16:45
Customized Attack Detection Under Power Industrial Control System

With the rapid development of information technology, the power system already has the typical characteristics of the information physical fusion system. The power industrial control system is widely used in the power industry. While improving the eciency, the economic bene ts have also been greatly improved. However, the dependence on information technology has also increased the vulnerability to malicious attacks. Power industry control system is facing a more serious threat. In this paper, we combine anomaly detection and data dimensionality reduction to propose a feature extraction method for iForest power measurement data, which not only ensures the targeting of attack detection in the data processing stage, but also takes into account the data quality of feature extraction. In addition, we use deep learning techniques to identify attack behavior characteristics and use captured features to detect attack behavior in real time. We prove the availability of the method through simulation of the IEEE 118-bus power systems.
Authors: Bin Wang (State Grid XinJiang Electric Power co LTD, Electric Power Research Institute), Ling He (State Grid XinJiang Electric Power co LTD, Electric Power Research Institute), Huiting Yang (State Grid XinJiang Electric Power co LTD, Electric Power Research Institute), Feng Li (State Grid XinJiang Electric Power co LTD, Electric Power Research Institute), Jie Fan (Global Enegry Interconnection Research Institute co. LTD),
Hide Authors & Abstract

Show Authors & Abstract
Room #2

Technical Sessions 6 13:30 - 15:00

Meeting Room: West, Session chaired by Prof. Lun-Ping Hung
13:30 - 13:45
An OFDMA-based neighbor discovery and tracking protocol for directional Ad-Hoc Network

In order to implement an effective Directed Medium Access Control Protocol and Routing Protocol (DMAC) in a directed ad hoc network, the node in the network should know the state of the neighboring nodes. However, due to the strong directivity of the directional antenna, it is difficult to sense signals in other directions, which causes problems such as link collision and node movement. In order to solve the above problems, this paper proposes a protocol based on Orthogonal Frequency Division Multiple Access (OFDMA) for directional neighbor discovery and tracking, and discusses the neighboring state partitioning method based on discovery time and multiple Resource Unit (RU) access and frame format design based on OFDMA. The simulation results show that compared with the Directional Transmission and Reception Algorithms protocol (DTRA), the efficiency of neighboring nodes in the protocol is 100\%, and the simulation results show that neighbor discovery and tracking improve network efficiency and save network resources.
Authors: Xiaojiao Hu (School of Electronics and Information, Northwestern Polytechnical University), Qi Yang (School of Electronics and Information, Northwestern Polytechnical University), Bo Li (School of Electronics and Information, Northwestern Polytechnical University), Zhongjiang Yan (Northwestern Polytechnical University), Mao Yang (School of Electronics and Information, Northwestern Polytechnical University),
Hide Authors & Abstract

Show Authors & Abstract
13:45 - 14:00
A combined routing path and node importance network invulnerability evaluating method for Ad Hoc network

In this paper, a new network invulnerability evaluation model based on routing path and node importance is proposed. The core of the invulnerability algorithm lies in two levels, comprehensive consideration of the influencing factors of network to point-to-line network invulnerability. The algorithm considers the influence of the proportion of important nodes of the network on the invulnerability of the network, and considers the influence of the number of paths between the communication nodes on the invulnerability of the path. Through simulation comparison, it is found that this algorithm improves the sensitivity of network invulnerability to the number of paths, and is suitable for communication networks in the case of multiple routing paths.
Authors: Weiling Zhou (School of Electronics and Information, Northwestern Polytechnical University), Bo Li (School of Electronics and Information, Northwestern Polytechnical University), Zhongjiang Yan (Northwestern Polytechnical University), Mao Yang (School of Electronics and Information, Northwestern Polytechnical University),
Hide Authors & Abstract

Show Authors & Abstract
14:00 - 14:15
Multi-BSS Association and Cooperation based Handoff Scheme for the Next Generation mmWave WiFi: IEEE 802.11ay

Wireless LAN based on IEEE 802.11 protocol standard is widely used due to its advantages of low cost, fast speed, flexibility and convenience. Among them, the coverage distance of the WLAN based on IEEE 802.11ay protocol standard is relatively short for the low frequency. In order to meet the needs of VR, HD video and other emerging services, the high frequency WLAN is highly intensive deployment. The movement of nodes is bound to cause multi-BSS handoff. WLAN handoff process is complex and consumes a lot of network signaling and time. Based on the advantages of multi-BSS association and cooperation, this paper designs a handoff protocol for high-frequency WiFi to complete multi-BSS handoff without interruption of business continuity. This paper introduces media access control layer(MAC) of IEEE 802.11ay standard protocol and AP Clustering based on it, and designs it to have multi-BSS handoff function. Through simulation verification and comparison with other multi-BSS handoff technologies, this protocol improves the throughput and reduces the time delay.
Authors: Yue Li (Northwestern Polytechnical University), Ping Zhao (Northwestern Polytechnical University), Bo Li (Northwestern Polytechnical University), Mao Yang (School of Electronics and Information, Northwestern Polytechnical University), Zhongjiang Yan (Northwestern Polytechnical University),
Hide Authors & Abstract

Show Authors & Abstract
14:15 - 14:30
Traffic Load Perception based OFDMA MAC Protocol for the Next Generation WLAN

Abstract-With the rapid development of WLAN and the proliferation of intelligent terminals, the current WLAN protocol is no longer able to meet the needs of users. Therefore, the next generation of wireless networks has emerged to meet the growing demand for user traffic. Orthogonal Frequency Division Multiple Access (OFDMA), which enables simultaneous transmission of data by different User Equipment (UEs), is considered to be one of the key technologies of 802.11ax. In order to achieve high throughput rates and low access latency to ensure quality of service (QoS), 802.11ax supports two uplink access modes: scheduling access and random access. However, how to adaptively and efficiently switch these two access mechanisms in the process of real-time operation of the system, and effectively reduce the drawbacks caused by these two mechanisms is a thorny problem, and there is little research in this direction. This paper proposes an evaluation mechanism of network traffic load based on OFDMA-MAC protocol, and its performance is verified by simulation. The simulation results show that the traffic load assessment mechanism effectively improves the network throughput and quality of service (QoS), and also adapts to the dynamic changes in network traffic.
Authors: Jianfei Cheng (Northwestern Polytechnical University), Bo Li (Northwestern Polytechnical University), Mao Yang (School of Electronics and Information, Northwestern Polytechnical University), Zhongjiang Yan (Northwestern Polytechnical University),
Hide Authors & Abstract

Show Authors & Abstract
14:30 - 14:45
Environment Sensing based Adaptive Acknowledgement and Backoff for the Next Generation WLAN

With the rapid development of wireless LAN technology in recent ten years, generations of wireless standards have been issued, and the widely discussed IEEE 802.11ax standards will be released in recent years, in which the concept of spatial reuse has been unanimously recognized by the industry. This phenomenon indicates that the development direction of wireless scenarios in the next decade is bound to be high-density, ultra-high-density deployment of cells, only such cell can meet people's rapidly increasing business needs for wireless networks. So for the next generation WLAN, how to efficiently provide reliable services for users in high-density deployment scenarios will be the most important problem to be solved. To solve this problem, this paper proposes ESBLA (Environment sensing based link adaptation) algorithm from the point of view of link adaptation. It effectively combines environment intelligent sensing and MAC layer transmission parameter adjustment. It not only solves the most difficult real-time problem in link adaptive algorithm, but also makes the adjustment of transmission parameter more reasonable according to the diversity of environment. The simulation results show that the algorithm can reduce the impact of intensive deployment interference as much as possible while guaranteeing high throughput.
Authors: Yuan Yan (Northwestern Polytechnical University), Bo Li (Northwestern Polytechnical University), Mao Yang (School of Electronics and Information, Northwestern Polytechnical University), Zhongjiang Yan (Northwestern Polytechnical University),
Hide Authors & Abstract

Show Authors & Abstract