Advances in Intelligent Systems and Computing 895

Ching-Nung YangSheng-Lung PengLakhmi C. Jain Editors

Security with Intelligent Computing and Big-data ServicesProceedings of the Second International Conference on Security with Intelligent Computing and Big Data Services (SICBS-2018)

Advances in Intelligent Systems and Computing

Volume 895

Series Editor

Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences,Warsaw, Poland

Advisory Editors

Nikhil R. Pal, Indian Statistical Institute, Kolkata, IndiaRafael Bello Perez, Faculty of Mathematics, Physics and Computing,Universidad Central de Las Villas, Santa Clara, CubaEmilio S. Corchado, University of Salamanca, Salamanca, SpainHani Hagras, Electronic Engineering, University of Essex, Colchester, UKLászló T. Kóczy, Department of Automation, Széchenyi István University,Gyor, HungaryVladik Kreinovich, Department of Computer Science, University of Texasat El Paso, El Paso, TX, USAChin-Teng Lin, Department of Electrical Engineering, National ChiaoTung University, Hsinchu, TaiwanJie Lu, Faculty of Engineering and Information Technology,University of Technology Sydney, Sydney, NSW, AustraliaPatricia Melin, Graduate Program of Computer Science, Tijuana Instituteof Technology, Tijuana, MexicoNadia Nedjah, Department of Electronics Engineering, University of Rio de Janeiro,Rio de Janeiro, BrazilNgoc Thanh Nguyen, Faculty of Computer Science and Management,Wrocław University of Technology, Wrocław, PolandJun Wang, Department of Mechanical and Automation Engineering,The Chinese University of Hong Kong, Shatin, Hong Kong

The series “Advances in Intelligent Systems and Computing” contains publicationson theory, applications, and design methods of Intelligent Systems and IntelligentComputing. Virtually all disciplines such as engineering, natural sciences, computerand information science, ICT, economics, business, e-commerce, environment,healthcare, life science are covered. The list of topics spans all the areas of modernintelligent systems and computing such as: computational intelligence, soft comput-ing including neural networks, fuzzy systems, evolutionary computing and the fusionof these paradigms, social intelligence, ambient intelligence, computational neuro-science, artificial life, virtual worlds and society, cognitive science and systems,Perception and Vision, DNA and immune based systems, self-organizing andadaptive systems, e-Learning and teaching, human-centered and human-centriccomputing, recommender systems, intelligent control, robotics and mechatronicsincluding human-machine teaming, knowledge-based paradigms, learning para-digms, machine ethics, intelligent data analysis, knowledge management, intelligentagents, intelligent decision making and support, intelligent network security, trustmanagement, interactive entertainment, Web intelligence and multimedia.

The publications within “Advances in Intelligent Systems and Computing” areprimarily proceedings of important conferences, symposia and congresses. Theycover significant recent developments in the field, both of a foundational andapplicable character. An important characteristic feature of the series is the shortpublication time and world-wide distribution. This permits a rapid and broaddissemination of research results.

** Indexing: The books of this series are submitted to ISI Proceedings,EI-Compendex, DBLP, SCOPUS, Google Scholar and Springerlink **

More information about this series at http://www.springer.com/series/11156

Ching-Nung Yang • Sheng-Lung Peng •

Lakhmi C. JainEditors

Security with IntelligentComputing and Big-dataServicesProceedings of the Second InternationalConference on Security with IntelligentComputing and Big Data Services(SICBS-2018)

123

EditorsChing-Nung YangDepartment of Computer Scienceand Information EngineeringNational Dong Hwa UniversityHualien, Taiwan

Sheng-Lung PengDepartment of Computer Scienceand Information EngineeringNational Dong Hwa UniversityHualien, Taiwan

Lakhmi C. JainUniversity of CanberraCanberra, ACT, Australia

University of Technology SydneyBroadway, Australia

Liverpool Hope UniversityLiverpool, UK

KES InternationalShoreham-by-sea, UK

ISSN 2194-5357 ISSN 2194-5365 (electronic)Advances in Intelligent Systems and ComputingISBN 978-3-030-16945-9 ISBN 978-3-030-16946-6 (eBook)https://doi.org/10.1007/978-3-030-16946-6

© Springer Nature Switzerland AG 2020This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or partof the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmissionor information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilarmethodology now known or hereafter developed.The use of general descriptive names, registered names, trademarks, service marks, etc. in thispublication does not imply, even in the absence of a specific statement, that such names are exempt fromthe relevant protective laws and regulations and therefore free for general use.The publisher, the authors and the editors are safe to assume that the advice and information in thisbook are believed to be true and accurate at the date of publication. Neither the publisher nor theauthors or the editors give a warranty, expressed or implied, with respect to the material containedherein or for any errors or omissions that may have been made. The publisher remains neutral with regardto jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AGThe registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

With the proliferation of security with intelligent computing and big-data services,the issues of information security, big data, intelligent computing, blockchaintechnology, and network security have attracted a growing number of researchers.The purpose of 2018 International Conference on Security with IntelligentComputing and Big-data Services (SICBS 2018 for short) is to provide a platformfor researchers, engineers, academicians, and industrial professionals from all overthe world to present their research results in security-related areas.

SICBS 2018 was held at Guilin, China, from December 14 to 16, 2018. Theconference brought together researchers from all regions around the world workingon a variety of fields and provided a stimulating forum for them to exchange ideasand report on their researches. The proceedings of SICBS 2018 consists of 75selected papers which were submitted to the conferences and peer-reviewed byconference committee members and international reviewers. All these papers coverthe following topics: Blockchain Technology and Application, MultimediaSecurity, Information Processing, Security of Network, Cloud and IoT,Cryptography and Cryptosystem, Learning and Intelligent Computing, andInformation Hiding.

On behalf of sponsors and conference committees, we would like to express ourgratitude to all of the authors, the reviewers, and the attendees for their contribu-tions and participation for SICBS 2018. Their high competence and expertiseknowledge enable us to prepare this high-quality program and make the confer-ences successful. Finally, we would like to wish you have good success in yourpresentations and social networking. Your strong supports are critical to the successof this conference. In addition, we are grateful to Springer for publishing theconference proceeding.

Finally, we would like to wish you have good success in your presentations andsocial networking. Your strong supports are critical to the success of this confer-ence. We hope that the participants will not only enjoy the technical program in

v

conference but also discover many beautiful places in Guilin, which has China’smost beautiful karst landscapes Li River and Yangshuo. Wishing you a fruitful andenjoyable SICBS 2018.

Ching-Nung YangSheng-Lung PengLakhmi C. Jain

vi Preface

Organization

2018 International Conference on Security with Intelligent Computingand Big-data Services (SICBS 2018)

Honorary Chairs

Han-Chieh Chao National Dong Hwa University, TaiwanLakhmi C. Jain University of Canberra, Australia,

University of Technology Sydney, Australia,Liverpool Hope University, UK,and KES International, UK

General Chairs

Ching-Nung Yang National Dong Hwa University, TaiwanMingzhe Liu Chengdu University of Technology, China

Organizing Chairs

Yi-Ning Liu Guilin University of Electronic Technology, ChinaRushi Lan Guilin University of Electronic Technology, China

Program Chairs

Sheng-Lung Peng National Dong Hwa University, TaiwanShiuh-Jeng Wang Central Police University, Taiwan

vii

Technical Program Committee

Jinsuk Baek Winston-Salem State University, USAZhifeng Bao RMIT University, AustraliaJalel Ben-Othman Université Paris 13, FranceNadia Bennani University of Lyon, FranceTao-Ku Chang National Dong Hwa University, TaiwanChao Chen Xidian University, ChinaChia-Mei Chen National Sun Yat-sen University, TaiwanJiageng Chen Central China Normal University, ChinaTzung-Her Chen National Chiayi University, TaiwanYu-Chi Chen Yuan Ze University, TaiwanBo-Chao Cheng National Chung Cheng University, TaiwanHung-Yu Chien National Chi Nan University, TaiwanIsao Echizen National Institute of Informatics, JapanAtilla Elci Aksaray University, TurkeyChun-I Fan National Sun Yat-sen University, TaiwanZhangjie Fu Nanjing University of Information Science

and Technology, ChinaClemente Galdi University of Napoli, ItalyAndy Guo University of Chinese Academy of Sciences, ChinaLein Harn University of Missouri—Kansas City, USAJi-Han Jiang National Formosa University, TaiwanCheonshik Kim Sejong University, KoreaWen-Chung Kuo National Yunlin University of Science and Technology,

TaiwanNarn-Yih Lee Southern Taiwan University of Science

and Technology, TaiwanZne-Jung Lee Huafan University, TaiwanMin Lei Beijing University of Posts and Telecommunications,

ChinaLi Li Hangzhou Dianzi University, ChinaPeng Li North China Electric Power University, ChinaChanglu Lin Fujian Normal University, ChinaFeng Liu University of Chinese Academy of Sciences, ChinaQi Liu Nanjing University of Information Science

and Technology, ChinaQuansheng Liu Université Bretagne Sud, FranceWen-Jie Liu Nanjing University of Information Science

and Technology, ChinaYanxiao Liu Xi’an University of Technology, ChinaDer-Chyuan Lou Chang Gung University, TaiwanWei Lu Sun Yat-sen University, ChinaJiafa Mao Zhejiang University of Technology, China

viii Organization

Chuan Qin University of Shanghai for Science and Technology,China

Maryam Sepehri University of Milan, ItalyHossain Shahriar Kennesaw State University, USAJian Shen Nanjing University of Information Science

and Technology, ChinaDongkyoo Shin Sejong University, KoreaHung-Min Sun National Tsing Hua University, TaiwanDan Tang Chengdu University of Information Technology, ChinaXiaohu Tang Southwest Jiaotong University, ChinaRay-Lin Tso National Chengchi University, TaiwanDaoshun Wang Tsinghua University, ChinaJiabao Wang Army Engineering University of PLA, ChinaJian Wang Nanjing University of Aeronautics and Astronautics,

ChinaShiow-Yang Wu National Dong Hwa University, TaiwanTzong-Sun Wu National Taiwan Ocean University, TaiwanXiaotian Wu Jinan University, ChinaZhe Xia Wuhan University of Technology, ChinaYasushi Yamaguchi University of Tokyo, JapanXuehu Yan National University of Defense Technology, ChinaWu-Chuan Yang I-Shou University, TaiwanKuo-Hui Yeh National Dong Hwa University, TaiwanXinpeng Zhang Fudan University, ChinaHong Zhao Fairleigh Dickinson University, USAZhili Zhou Nanjing University of Information Science

and Technology, China

Organization ix

Contents

Blockchain Technology and Application

Fast Adaptive Blockchain’s Consensus Algorithm via WlanMesh Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Xin Jiang, Mingzhe Liu, Feixiang Zhao, Qin Zhou, and Ruili Wang

Toward a Blockchain Based Image Network Copyright TransactionProtection Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Chengqiang Zhao, Mingzhe Liu, Yanhan Yang, Feixiang Zhao,and Shijie Chen

A Blockchain-Based Scheme for Secure Sharing of X-RayMedical Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Bingqi Liu, Mingzhe Liu, Xin Jiang, Feixiang Zhao, and Ruili Wang

Blockchain Based Data Trust Sharing Mechanismin the Supply Chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43Luya Wang and Shaoyong Guo

Designing Smart-Contract Based Auctions . . . . . . . . . . . . . . . . . . . . . . . 54Chiara Braghin, Stelvio Cimato, Ernesto Damiani,and Michael Baronchelli

Blockchain-Based Autonomous Peer to Peer Information InteractionSystem in Financial Audit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65Yu Zhuo, Dong Aiqiang, and Yan Yong

Multimedia Security

Relations Between Secret Sharing and Secret Image Sharing . . . . . . . . . 79Xuehu Yan, Jinming Li, Yuliang Lu, Lintao Liu, Guozheng Yang,and Huixian Chen

xi

Enhanced Secret Image Sharing Using POB Number System . . . . . . . . 94Yan-Xiao Liu and Ching-Nung Yang

Study on Security Enhancing of Generalized Exploiting ModificationDirections in Data Hiding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103Wen-Chung Kuo, Ren-Jun Xiao, Chun-Cheng Wang, and Yu-Chih Huang

High-Fidelity Reversible Data Hiding in JPEG Images Basedon Two-Dimensional Histogram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116Tengfei Dong, Zhigao Hong, and Zhaoxia Yin

A Lossless Polynomial-Based Secret Image Sharing Scheme Utilizingthe Filtering Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129Lintao Liu, Yuliang Lu, Xuehu Yan, Wanmeng Ding, and Qitian Xuan

CT Image Secret Sharing Based on Chaotic Map and SingularValue Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140Feixiang Zhao, Mingzhe Liu, Xianghe Liu, Xin Jiang, and Zhirong Tang

Information Processing

Robust Speaker Recognition Using Improved GFCC and AdaptiveFeature Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159Xingyu Zhang, Xia Zou, Meng Sun, and Penglong Wu

Outlier Detection Based on Local Density of Vector Dot Productin Data Stream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170Zhaoyu Shou, Fengbo Zou, Hao Tian, and Simin Li

Multi-level Competitive Swarm Optimizer for Large ScaleOptimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185Li Zhang, Yu Zhu, Si Zhong, Rushi Lan, and Xiaonan Luo

Experimental Comparison of Free IP Geolocation Services . . . . . . . . . . 198Wei Xu, Yaodong Tao, and Xin Guan

Integrating Multiple Feature Descriptors for Computed TomographyImage Retrieval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209Xiaoqin Wang, Huadeng Wang, Rushi Lan, and Xiaonan Luo

Automatic Forgery Localization via Artifacts Analysis of JPEGand Resampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221Hongbin Wei, Heng Yao, Chuan Qin, and Zhenjun Tang

Arbitrary Style Transfer of Facial Image Based on Feed-ForwardNetwork and Its Application in Aesthetic QR Code . . . . . . . . . . . . . . . . 235Shanqing Zhang, Shengqi Su, Li Li, Jianfeng Lu, Ching-Chun Chang,and Qili Zhou

xii Contents

Aesthetic QR Code Authentication Based on Directed PeriodicTexture Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251Li Li, Min He, Jier Yu, Jianfeng Lu, Qili Zhou, Xiaoqing Feng,and Chin-Chen Chang

A Heterogeneous Multiprocessor Independent Task SchedulingAlgorithm Based on Improved PSO . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267Xiaohui Cheng and Fei Dai

Cross-Domain Text Sentiment Classification Basedon Wasserstein Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280Guoyong Cai, Qiang Lin, and Nannan Chen

WebRTC-Based On-Site Operation and Maintenance AdaptiveVideo Streaming Rate Control Strategy . . . . . . . . . . . . . . . . . . . . . . . . . 292Chuang Liu, Sujie Shao, Shaoyong Guo, and Xuesong Qiu

Evaluation of Enzymatic Extract with Lipase Activity of YarrowiaLipolytica. An Application of Data Mining for the Food IndustryWastewater Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304Heidy Posso Mendoza, Rosangela Pérez Salinas, Arnulfo Tarón Dunoyer,Claudia Carvajal Tatis, W. B. Morgado-Gamero,Margarita Castillo Ramírez, and Alexander Parody

Load-Balancing-Based Reliable Mapping Algorithm for VirtualResources in Power Communication Network . . . . . . . . . . . . . . . . . . . . 314Nie Junhao, Qi Feng, Li Wenjing, and Zhang Zhe

Security of Network, Cloud and IoT

Enhancing Network Intrusion Detection System Method (NIDS) UsingMutual Information (RF-CIFE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329Nyiribakwe Dominique and Zhuo Ma

You Are What You Search: Attribute Inference Attacks ThroughWeb Search Queries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343Tianyu Du, Tao Tao, Bijing Liu, Xueqi Jin, Jinfeng Li, and Shouling Ji

New Publicly Verifiable Cloud Data Deletion Schemewith Efficient Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359Changsong Yang and Xiaoling Tao

A Transient Grid Security Control Algorithm Basedon EMS System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373Yang Su, Song Liu, Zhihong Liang, Zhizhong Qiao, and Xiaodong Li

Security Analysis of Bioinformatics WEB Application . . . . . . . . . . . . . . 383Tao Tao, Yuan Chen, Bijing Liu, Xueqi Jin, Mingyuan Yan,and Shouling Ji

Contents xiii

Dynamic Network Configuration: An Effective Defensive Protocolfor Public Blockchain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398Zhengwei Jiang, Chenyang Lv, Bo Zhang, Chao Zhang, Wei Lu,and Shouling Ji

Survival Model for WiFi Usage Forecasting in NationalFormosa University . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414Jutarat Kositnitikul and Ji-Han Jiang

An Auditing Scheme for Cloud-Based Checkout Systems . . . . . . . . . . . 423Tao-Ku Chang and Cheng-Yen Lu

Implementation of an IP Management and Risk Assessment SystemBased on PageRank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438Chia-Ling Hou, Cheng-Chung Kuo, I-Hsien Liu, and Chu-Sing Yang

Integrated Forensic Tool for Network Attacks . . . . . . . . . . . . . . . . . . . . 451Chia-Mei Chen, Gu-Hsin Lai, and Zheng-Xun Tsai

Cryptanalysis of Anonymous Three Factor-Based AuthenticationSchemes for Multi-server Environment . . . . . . . . . . . . . . . . . . . . . . . . . 456Jiaqing Mo, Hang Chen, and Wei Shen

LWE-Based Single-Server Block Private InformationRetrieval Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469Shuai Liu and Bin Hu

Design and Implementation of an Automatic Scanning Tool of SQLInjection Vulnerability Based on Web Crawler . . . . . . . . . . . . . . . . . . . 481Xiaochun Lei, Jiashi Qu, Gang Yao, Junyan Chen, and Xin Shen

Malware Detection Based on Opcode Sequence and ResNet . . . . . . . . . 489Xuetao Zhang, Meng Sun, Jiabao Wang, and Jinshuang Wang

Design of Elderly Care System Integrated with SLAM Algorithm . . . . . 503Jun-Yan Chen and Long Huang

A Fog-Based Collusion Detection System . . . . . . . . . . . . . . . . . . . . . . . . 514Po-Yang Hsiung, Chih Hung Li, Shih Hung Chang, and Bo-Chao Cheng

A Real-Time Online Security Situation Prediction Algorithmfor Power Network Based on Adaboost and SVM . . . . . . . . . . . . . . . . . 526Haizhu Wang, Wenxin Guo, Ruifeng Zhao, Bo Zhou, and Chao Hu

A Pi-Based Beehive IoT System Design . . . . . . . . . . . . . . . . . . . . . . . . . 535Yi-Liang Chen, Hung-Yu Chien, Ting-Hsuan Hsu, Yi-Jhen Jing,Chun-Yu Lin, and Yi-Chun Lin

PLVSNs: A Privacy-Preserving Location-Sharing Systemfor Vehicular Social Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544Chang Xu, Xuan Xie, Liehuang Zhu, Chuan Zhang, and Yining Liu

xiv Contents

Cryptography and Cryptosystem

Witness-Based Searchable Encryption with Aggregative Trapdoor . . . . 561Xin Xie, Yu-Chi Chen, Jun-Rui Wang, and Yingjie Wu

Robust/Recover Provable Data Possession Protocol . . . . . . . . . . . . . . . . 574Chao Feng, Honghong Wang, Wenbo Wan, Qinghua Li, and Fangzhou Xu

A Provably Secure Scalable Revocable Identity-Based SignatureScheme Without Bilinear Pairings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588Changji Wang, Hui Huang, and Yuan Yuan

Efficient Computation Method of Participants’ Weights in Shamir’sSecret Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598Long Li, Tianlong Gu, Liang Chang, and Jingjing Li

An Identity-Set-Based Provable Data Possession Scheme . . . . . . . . . . . . 606Changlu Lin, Fucai Luo, Jinglong Luo, and Yali Liu

(2, 2) Threshold Robust Visual Secret Sharing Scheme for QR CodeBased on Pad Codewords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619Longdan Tan, Yuliang Lu, Xuehu Yan, Lintao Liu, and Jinrui Chen

Hybrid Chain Based Hierarchical Name Resolution Service in NamedData Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 629Zhuo Yu, Aiqiang Dong, Xin Wei, Shaoyong Guo, and Yong Yan

A Secret Sharing Scheme of Dynamic Threshold with CheatersDetection and Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638Yong-Zhong He, Xue-Jiao Xu, Dao-Shun Wang, Shun-Dong Li,Xiang-Hui Zhao, and Ching-Nung Yang

Game-Based Security Proofs for Secret Sharing Schemes . . . . . . . . . . . 650Zhe Xia, Zhen Yang, Shengwu Xiong, and Ching-Fang Hsu

Dress Identification for Camp Security . . . . . . . . . . . . . . . . . . . . . . . . . 661Jiabao Wang, Yang Li, Yihang Xiong, Zhixuan Zhao, and Dexing Kong

ETSB: Energy Trading System Based on Blockchain . . . . . . . . . . . . . . 673Xia Dong, Wei Zaoyu, Mao Hua, Xu Jing, Yang Debo, Wang Fanjin,and Bi Wei

An Analysis on Inverted Mirrored Moiré and Ribbonof Band Moiré . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685Hui-Ying Liu, Dao-Shun Wang, Xiang-Hui Zhao, Shun-Dong Li,and Ching-Nung Yang

Contents xv

Learning and Intelligent Computing

A LSTM-Based Approach to Haze Prediction Using a Self-organizingSingle Hidden Layer Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701Xiaodong Liu, Qi Liu, Yanyun Zou, and Qiang Liu

Target Information Fusion Based on Memory Networkfor Aspect-Level Sentiment Classification . . . . . . . . . . . . . . . . . . . . . . . . 707Zhaochuan Wei, Jun Peng, Xiaodong Cai, and Guangming He

Dangerous Objects Detection of X-Ray Images Using ConvolutionNeural Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714Lekang Zou, Tanaka Yusuke, and Iba Hitoshi

An Efficient JPEG Steganalysis Model Based on Deep Learning . . . . . . 729Lin Gan, Yang Cheng, Yu Yang, Linfeng Shen, and Zhexuan Dong

Cross-Domain Image Steganography Based on GANs . . . . . . . . . . . . . . 743Yaojie Wang, Xiaoyuan Yang, and Jia Liu

Facial Expression Recognition by Transfer Learningfor Small Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756Jianjun Li, Siming Huang, Xin Zhang, Xiaofeng Fu, Ching-Chun Chang,Zhuo Tang, and Zhenxing Luo

A Network Intrusion Detection Model Based on ConvolutionalNeural Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771Wenwei Tao, Wenzhe Zhang, Chao Hu, and Chaohui Hu

A Dual-Branch CNN Structure for Deformable Object Detection . . . . . 784Jianjun Li, Kai Zheng, Xin Zhang, Zhenxing Luo, Zhuo Tang,Ching-Chun Chang, Yuqi Lin, and Peiqi Tang

Initial Sensitivity Optimization Algorithm for Fuzzy-C-MeansBased on Particle Swarm Optimization . . . . . . . . . . . . . . . . . . . . . . . . . 798Zilong Ye, Feng Qi, Jingquan Li, Yanjun Liu, and Han Su

Multi-modal Behavioral Information-Aware Recommendationwith Recurrent Neural Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808Guoyong Cai, Nannan Chen, Weidong Gu, and Jiao Pan

Energy-Efficient Task Caching and Offloading Strategy in MobileEdge Computing Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824Qian Chen, Zhoubin Liu, Linna Ruan, Zixiang Wang, Sujie Shao,and Feng Qi

Emotion Detection in Cross-Lingual Text Basedon Bidirectional LSTM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838Han Ren, Jing Wan, and Yafeng Ren

xvi Contents

Information Hiding

Sharing More Information in Visual Cryptography Schemewith Block Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849Peng Li, Ching-Nung Yang, Dan Tang, and Jianfeng Ma

High Capacity Data Hiding Based on AMBTC and Interpolation . . . . . 858Cheonshik Kim, Dongkyoo Shin, and Ching-Nung Yang

A Reversible Data Hiding Scheme in Encrypted Image with LSBCompression and MSB Replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . 868Wei Zhang, Qing Zhou, Zhenjun Tang, Heng Yao, and Chuan Qin

Halftone Image Steganography with Flippability MeasurementBased on Pixel Pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879Wanteng Liu, Wei Lu, Xiaolin Yin, Junhong Zhang, and Yuileong Yeung

Reversible Data Hiding in Binary Images by Symmetrical FlippingDegree Histogram Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891Xiaolin Yin, Wei Lu, Wanteng Liu, and Junhong Zhang

Binary Image Steganalysis Based on Local Residual Patterns . . . . . . . . 904Ruipeng Li, Lingwen Zeng, Wei Lu, and Junjia Chen

Robust Steganography Using Texture Synthesis Based on LBP . . . . . . . 914Weiyi Wei and Chengfeng A

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925

Contents xvii

Blockchain Technology and Application

Fast Adaptive Blockchain’s ConsensusAlgorithm via Wlan Mesh Network

Xin Jiang1, Mingzhe Liu1(B), Feixiang Zhao1, Qin Zhou1, and Ruili Wang2

1 State Key Laboratory of Geohazard Prevention and Geoenvironment Protection,Chengdu University of Technology, Chengdu, Sichuan, China

liumz@cdut.edu.cn2 Institute of Natural and Mathematical Sciences, Massey University,

Auckland, New Zealand

Abstract. This paper presents a decentralised and fast adaptive blockchain’s consensus algorithm with maximum voter privacy using wlanmesh network. The algorithm is suitable for consortium blockchain andprivate blockchain, and is written as a smart contract for HyperledgerFabric. Unlike previously proposed blockchain’s consensus protocols, thisis the first implementation that does not rely on any trusted authority tocompute the tally or to protect the voter’s privacy. Instead, the algorithmis a fast adaptive protocol, and each voter is in control of the privacy oftheir own vote such that it can only be breached by a full collusion involv-ing all other voters. The execution of the protocol is enforced using theconsensus mechanism that also secures the Fabric blockchain. This papertests the implementation on Fabric’s official test network to demonstrateits feasibility. Also, this paper provides a computational breakdown ofits execution cost.

Keywords: Consensus algorithm · Blockchain · Wlan mesh ·Hyperledger Fabric

1 Introduction

Hyperledger Fabric is the most popular cryptocurrency of blockchain as of 2018.It relies on the same innovation behind Bitcoin [1]: namely, the blockchain whichis an append-only ledger. The blockchain is maintained by a decentralised andopen-membership peer-to-peer network. The purpose of the blockchain was toremove the centralised role of banks for maintaining a financial ledger. Today,researchers are trying to re-use the Blockchain to solve further open problemssuch as coordinating the Internet of Things [2], carbon dating [3], and healthcare[4]. However, How to reach consensus fastly has become a bottleneck in thedevelopment of blockchain in above fields.

Supported by organization X.

c© Springer Nature Switzerland AG 2020C.-N. Yang et al. (Eds.): SICBS 2018, AISC 895, pp. 3–16, 2020.https://doi.org/10.1007/978-3-030-16946-6_1

4 X. Jiang et al.

In this paper, we focus on decentralised internet consensus of blockchainusing wlan mesh network. This algorithm is designed as a voting protocol. E-voting protocols that support verifiability normally assume the existence of apublic bulletin board that provides a consistent view to all voters. In practice,an example of implementing the public bulletin board can be seen in the yearlyelections of the International Association of Cryptologic Research (IACR) [5].They use the Helios voting system [6] whose bulletin board is implemented asa single web server. This server is trusted to provide a consistent view to allvoters. Instead of such a trust assumption, we explore the feasibility of usingthe Blockchain as a public bulletin board. Furthermore, we consider a decen-tralised election setting in which the voters are responsible for coordinating thecommunication amongst themselves. Thus, we also examine the suitability.

2 Related Works

In recent years, the unique advantages of blockchain technology have been paidattention to by academic circles. There are some research methods of blockchainin consensus algorithm. BBLAST system is recent one of them, it solved the prob-lem of low trust centralized ledger held by a single third party, high trust decen-tralized forms held by different entities, or in other words, verification nodes [7].

Gramoli [8] discussed the mainstream blockchain consistency algorithm andthe classic Byzantine consensus to re-examine the blockchain context. To againstbitcoin and the Ethum consensus algorithm, the document [9] worked to proveconsensus algorithm of mining dilemma in the process of analysising PoW (Proofof the work) strategy choice for the existence of Nash equilibrium. In general, thissolution came from the perspective of game theory, analyzed the PoW consensusalgorithm, and provided new ideas and methods for the further design of theconsensus algorithm based on game theory.

To overcome performance problem, document [10] proposed a public sup-ply chain system based on double-chain structure. That results of test showedthat the two-chain structure of supply chain based on agricultural product sup-ply chain could consider the openness and security of transaction informationas well as the privacy of enterprise information, and could achieve rent-seekingand matching adaptively. Ref. [11] proposed a new blockchain consistency algo-rithm, introduced the two-stage delivery and quorum vote process, and solvedthe legitimacy verification problem in decentralized environment by using thedistributed ledger feature of the blockchain protocol. Compared with the tra-ditional Byzantine consensus, the algorithm reduced the number of messagepassing, improved system fault tolerance. Artical [12] instantiated a provablysecure OKSA solution at transport stage, reduced a round interaction and con-stant communication cost. The analysis and evaluation showed that the searchchain could remain reasonably cost effective without loss of retrieval privacy.The literature [13] proposed a security certification scheme that it could managehuman-centric solutions (SAMS) to verify resource information in participatingmobile devices and processing in MRM resource pools.

Fast Adaptive Blockchain’s Consensus Algorithm via Wlan Mesh Network 5

In order to verify the SAMS of MRM, the data of FalsifICA was tested.The test results showed that data tampering was impossible. To achieve thefault tolerance of the XFT consensus algorithm, the solution proposed by thedocument [14] presented a Byzantine consensus algorithm based on the Gossipprotocol, which could enable the system to tolerate less than half of the nodesas byzantines. At the same time, the system has better scalability, and it ishelpful to identify malicious nodes for correct nodes in the blockchain systemsince the unified data structure is adopted. In order to improve operation effi-ciency of the consensus algorithm in blockchain system, the document [15] firstlyintroduced various potential of blocks in the chain of consensus algorithm opti-mization scheme, then optimized dBFT consensus process by combining withaggregate signature technology and bilinear mapping technology. The optimizedaggregate dBFT consensus algorithm can effectively reduce the space complexityof signature in blockchain system. The document [16] also explored how differ-ent network conditions can change the results of consistency between nodes.Besides, Bach’s analysis [17] focused on the algorithm steps: the scalability algo-rithm adopted by each consistent algorithm, and the time and security risks ofthe algorithm reward validator verification block exist in the algorithm.

As above of all, generalizability is a basic problem in distributed computing,although the question is known that existing agreements have been designed overthe past three years to address consensus. Under various assumptions with theadvent of chains, various consensuses have come to high performance. However,the Byzantine fault tolerance (BFT) protocol is not suitable. For a high num-ber of participants, it must be accommodated. Regarding to performance andsecurity, one can use novel network techniques and trusted computing [18,19].

3 Method

3.1 Consensus Algorithm

The consensus algorithm is a decentralized two-round protocol designed for sup-porting small-scale boardroom voting. In the first round, all voters register theirintention to vote in the election, and in the second round, all voters cast theirvote. The systems assumes an authenticated broadcast channel is available to allvoters. The self-tallying property allows anyone (including non-voters) to com-pute the tally after observing messages from the other voters. In this paper, weonly consider an election with two options, e.g., yes/no. Extending to multiplevoting options, and a security proof of the protocol can be found in [20].

A description of the Open Vote Network is as follows. First, all n votersagree on (G, g) where G denotes a finite cyclic group of prime order q n whichthe Decisional Diffie-Hellman (DDH) problem is intractable, and g is a generatorin G. A list of eligible voters (P1, P2, ..., Pn) is established and each eligible voterPi selects a random value as their private voting key.

Round 1. Every voter Pi broadcasts their voting key gxi and a (non-interactive) zero knowledge proof ZMP (xi) to prove knowledge of the exponent

6 X. Jiang et al.

xi on the public bulletin board. ZMP (xi) is implemented as a Schnorr proof[21] made non-interactive using the Fiat Shamir heuristic [22].

At the end, all voters check the validity of all zero knowledge proofs beforecomputing a list of reconstructed keys:

Yi =i−1∏

j=1

gxj/

n∏

j=i+1

gxj (1)

Implicitly setting Yi = gyi , the above calculation ensures that∑

i xiyi = 0.

Round 2. Every voter broadcasts gxiyigvi and a (non-interactive) zero knowl-edge proof to prove that vi is either no or yes (with respect to 0 or 1) vote. Thisone-out-of-two zero knowledge proof is implemented using the Cramer, Damg̊ardand Schoenmakers (CDS) technique.

All zero knowledge proofs must be verified before computing the tally toensure the encrypted votes are well-formed. Once the final vote has been cast,then anyone (including non-voters) can compute

∏i g

xiyigvi and calculate g∑

i vi

since∏

i gxiyi = 1. The discrete logarithm of g

∑i vi is bounded by the number

of voters and is a relatively small value. Hence the tally of yes votes can becalculated subsequently by exhaustive search.

Note that for the election tally to be computable, all the voters who havebroadcast their voting key in Round 1 must broadcast their encrypted vote inRound 2. Also note that in Round 2, the last voter to publish their encryptedvote has the ability to compute the tally before broadcasting their encrypted vote(by simulating that he would send a no-vote). Depending on the computed tally,he may change his vote choice. In our implementation, we address this issue byrequiring all voters to commit to their votes before revealing them, which addsanother round of commitment to the protocol.

The decentralised nature of the consensus algorithm makes it suitable toimplement over wlan mesh network. Wlan mesh could be used as the privateblockchain to store the voting data for the consensus algorithm.

3.2 Structure of Implementation

There are two smart contracts that are both written in Fabric’s Solidity lan-guage. The first contract is called the voting contract. It implements the votingprotocol, controls the election process and verifies the two types of zero knowl-edge proofs we have in the Open Vote Network. The second contract is calledthe cryptography contract. It distributes the code for creating the two types ofzero knowledge proofs 3. This provides all voters with the same cryptographycode that can be used locally without interacting with the Ethereum network.We have also provided three HTML5/JavaScript pages for the users, as shown inthe Fig. 1. We assume that voters and the election administrator have their ownFabric accounts. The Web3 framework is provided by the Ethereum Foundationto faciltiate communication between a user’s web browser and their Fabric client.The user can unlock their Fabric account (decrypt their Fabric private key using

Fast Adaptive Blockchain’s Consensus Algorithm via Wlan Mesh Network 7

a password) and authorise transactions directly from the web browser. There isno need for the user to interact with an Ethereum wallet, and the Fabric clientcan run in the background as a daemon.

Fig. 1. There are five stages to the election.

Election Administrator. This includes establishing the list of eligible voters,setting the election question, and activating a list of timers to ensure the electionprogresses in a timely manner. The latter includes notifying Ethereum to beginregistration, to close registration and begin the election, and to close voting andcompute the tally.

Voter. Voter can register for an election, and once registered must cast theirvote.

Observer. Observer can watch the election’s progress consisting of the electionadministrator starting and closing each stage and voters registering and castingvotes. The running tally is not computable.

3.3 Setup Flow

We defined 2 typical user setup scenarios: (1) New nodes – How to privatize yournodes first time. (2) Privatized nodes – How to re-privatize your nodes. Whenfirst open one node never been privatized, our nodes will be in public mesh state,user can use phone APP to privatize the devices and setup AP configurations.Figure 2 is an example showing the new devices setup sequence: Fig. 3 is anexample showing a new device how to join a privatized group:

Definitions, Acronyms and Abbreviations:

WMD: WIFI manager daemon.Master: The device which is in private mesh network and connected to AP.Slave: The device which is in private mesh network.Public Mesh: Mesh network with all known mesh ID.Private Mesh: Mesh network with privatized mesh ID.

8 X. Jiang et al.

Fig. 2. Flow of new nodes.

If there is a privatized device want to join another privatized group, what we needto do is press reset button to clear privatized information and reboot. Figure 4is an example showing a privatized device how to join another privatized group.

3.4 Node Discovery Protocol

We use SSDP protocol to discovery node, the structure of SSDP is designedas Fig. 5. According to the protocol, when a control point (client) accesses thenetwork, it can send “ssdp: discover” messages to a SSDP port with a specificmulticast address using the M-SEARCH method. When the device listens formessages sent by the control point on the reserved multicast address, it analyzesthe service requested by the control point. If it provides the service requested bythe control point itself, the device will directly respond to the request by unicast,see Fig. 6.

Fast Adaptive Blockchain’s Consensus Algorithm via Wlan Mesh Network 9

10 X. Jiang et al.

Fig. 4. Flow of finding privatized nodes.

Fig. 5. Internal structure of SSDP.

State: Pre-Role, Role is “slave without master information” (could be slaverole but without master parameters values.); Send broadcast “request masterinformation with slave role”; Definer responses the broadcast to get the AP RSSIvalue; Definer responses the broadcast to set it to master role, then connectthe AP and change to “Master” state. If connect AP failed, it will transferto “Definer” state. Master or slave responses the broadcast to set the master

Fast Adaptive Blockchain’s Consensus Algorithm via Wlan Mesh Network 11

Fig. 6. Setup flow of SSDP.

information, then save master information and change to “Slave” state. If thereis no change on the state after 8 s timeout, will transfer back to “No Role” state.

State: Definer, Role is definer (or slave but send “definer” broadcast); Sendbroadcast “definer”; Receive “request master information with slave role” to getthe AP RSSI value from the broadcaster; Receive “Definer” broadcast, checkIP (or UUID), if small than self’s, then change to “Pre-Role” state; After 2 stimeout, set the highest RSSI slave to the master and change to “Pre-Role” state(or change to “Slave” state), or change self to the “Master” state if self is thehighest RSSI one or no slave found; Master or slave responses the broadcast toset the master information, then save master information and change to “Slave”state.

State: Slave. Role is slave with master information saved; Receive “requestmaster information without role/with slave role” and “definer” broadcast, setthe master information to the broadcaster; Receive master’s heartbeat packet.

12 X. Jiang et al.

Fig. 7. Master vote strategy.

State: Master. Role is master with master information saved; Receive“request master information without role/with slave role” and “definer” broad-cast, set the master information to the broadcaster; Send out heartbeat packet.

Special case: When master leave actively, master set new master and sendto neighbors’ the change; When master leave suddenly, slave has timeout andchange to “No Role” state. Only neighbor slaves monitor the master’s heartbeatpackets. If master leaves suddenly, only neighbor slaves change to “No Role”state. Next jump slaves will be set to “Pre-Role” by neighbor slaves or nextjump slaves.

4 Experiments and Evaluation

4.1 Experiments

Our implementation was deployed on Fabric’s official test network that canmimic the production network. We sent 126 transactions to simulate forty votersparticipating in the protocol. Each transaction’s computational cost is outlined

Fast Adaptive Blockchain’s Consensus Algorithm via Wlan Mesh Network 13

5 10 15 20 25 30 35 40 45 50 55 600.0

0.5

1.0

1.5

2.0

2.5C

ost(s

)

Number of voters

Election administrator

Fig. 8. The time cost for the election administrator based on the number of votersparticipating in the election.

in Table 1. Each transaction by the election administrator is broadcast onlyonce, and each transaction by a voter is broadcast once per voter, i.e., a total of40 times.

Figure 8 demonstrates the results of our experiment and highlights the break-down of the election administrator’s time consumption. Except for opening reg-istration, the time cost for each task increases linearly with the number of voters.The time limit for a block was set at 4.7 s by the miners before the recent DoSattacks. This means that the smart contract reaches the computation and storagelimit if it is computing the voter’s reconstructed keys for around sixty registeredvoters. This limit exists as all keys are computed in a single transaction and thetime used must be less than the block’s time limit. To avoid reaching this blocklimit, we currently recommend a safe upper limit of around 50 voters. However,the contract can be modified to perform the processing in batches and allowmultiple transactions to complete the task.

14 X. Jiang et al.

4.2 Evaluation

Table 1 outlines the timing analysis measurements for tasks in the consensusalgorithm. All time measurements are rounded up to the next whole millisecond.We use the Web3 framework to facilitate communication between the webbrowser and the Fabric daemon. All tasks are executed using .call() that allowsus to measure the code’s computation time on the local daemon.

The wlan mesh is responsible for creating the proofs for the voter. The timerequired to create the proofs is 81 ms for the Schnorr proof and 461 ms for theone-out-of-two zero knowledge proof. These actions are always executed using.call() as this contract should never receive transactions.

The voting smart contract is responsible for enforcing the election process.Registering a vote involves verifying the Schnorr zero knowledge proof and intotal requires 142 ms. To begin the election requires computing the reconstructedpublic keys which takes 277 ms in total for forty voters. Casting a vote involvesverifying the one-out-of-two zero knowledge proof which requires 573 ms. Tally-ing involves summing all cast votes and brute-forcing the discrete logarithm ofthe result and on average takes around 132 ms.

Table 1. A time analysis for actions that run on the Fabric daemon.

Action Avg. Time (ms)

Create ZKP(x) 81

Register voting key 142

Begin election 277

Create 1-out-of-2 ZKP 461

Cast vote 573

Tally 132

5 Conclusion

In this paper, we have presented a fast adaptive blockchain consensus algorithmimplementation by using aximum voter privacy that runs on Fabric. Our imple-mentation was tested on the official Fabric test network with forty simulatedvoters. We have shown that our implementation can be readily used with mini-mal setup for elections at a little cost per voter. The cost can be considered rea-sonable as this voting protocol provides maximum voter privacy and is publiclyverifiable. This is the first implementation of a decentralised internet consensusprotocol running on wlan mesh network. It uses the Fabric blockchain not justas a public bulletin board, but more importantly, as a platform for consensuscomputing that enforces the correct execution of the voting protocol.

In future work, we will investigate the feasibility of running a national-scaleelection over the blockchain. Based on the knowledge gained from this paper, we