Parallel Filtering-Communication mode: 网络通讯/融合与节点滤波计算同步进行 网络协同“一致性”研究如火如荼十余载，源起复杂网络控制，后发展至分布式滤波和跟踪。但是现有所有的分布式滤波基本都是“滤波-融合-再滤波-再融合”这种“你方唱罢我登台，轮番上阵”的串行方式，或者说“鸡生蛋，蛋孵鸡”这种相互依赖迭代模式：所融合的对象就是滤波结果，而下一轮滤波的先验就是融合结果。 0-1的突破： 提出了 “一边滤波一边通讯融合”并行模式（即网络通讯融合与节点滤波计算同步进行， Parallel Filtering-Communication mode ） ，难点在于：没有鸡(滤波)怎么来的蛋(融合)？没有蛋(融合)又怎么来的鸡(滤波)？这就是我们Engineer们 施展拳脚 了。。。 团队近期两篇论文 （全世界独此两篇?） ： 第1篇基于粒子滤波，借助Importance Sampling方法，巧妙实现了 高度 Parallel Filtering-Communication ； 第2篇基于高斯混合GM滤波， Parallel Filtering-Communication 实现难度较大，所实现的融合对象仅仅是目标数cardinality的估计，融合层次教浅有待进一步研究。 两个工作分别对最具代表性的两类滤波后验近似形式进行了“滤波与通信并行”机制设计，可以推广到诸多以粒子滤波和高斯混合为基底的其他多传感实时滤波器设计。 网络通讯与节点滤波计算同步进行的优势自不必说， 甚至有些场景下是唯一选择 ！比 如局部节点的滤波计算占据了整个传感器扫描周期，根本没有剩余时间去搞通讯和信息融合 ---- 而这将伴随着传感器扫描周期越来越快变得普遍。。即使有些剩余时间，因为平行通讯可以完成更多周期的信息交换，传播更远，网络收益更大。。 1 A Distributed Particle-PHD Filter with Arithmetic-Average PHD Fusion Tiancheng Li , Franz Hlawatsch (Submitted on 17 Dec 2017 ( v1 ), last revised 20 Dec 2018 (this version, v2)) We propose a particle-based distributed PHD filter for tracking an unknown, time-varying number of targets. To reduce communication, the local PHD filters at neighboring sensors communicate Gaussian mixture (GM) parameters. In contrast to most existing distributed PHD filters, our filter employs an `arithmetic average' fusion. For particles--GM conversion, we use a method that avoids particle clustering and enables a significance-based pruning of the GM components. For GM--particles conversion, we develop an importance sampling based method that enables a parallelization of filtering and dissemination/fusion operations . The proposed distributed particle-PHD filter is able to integrate GM-based local PHD filters. Simulations demonstrate the excellent performance and small communication and computation requirements of our filter. Comments: 13 pages, codes available upon e-mail request Subjects: Systems and Control (cs.SY) ; Distributed, Parallel, and Cluster Computing (cs.DC) Cite as: arXiv:1712.06128 (or arXiv:1712.06128v2 for this version) submission Information Fusion，under revision 2. A Parallel Filtering-Communication based Cardinality Consensus Approach for Real-time Distributed PHD Filtering Tiancheng Li ; Mahendra Mallick ; Quan Pan Abstract: This paper proposes a new cardinality consensus (CC) approach called “prior-CC” to distributed probability hypothesis density (PHD) filtering based on a decentralized sensor network. In our approach, network-wide average consensus is sought with respect to the prior cardinality estimate. Unlike existing serial filtering-communication approaches, the prior-CC scheme allows the internode communication to be performed in parallel with the local filter calculation and requires only a small amount of interfacing fusion calculation and communication. This enables real-time filtering that minimizes data delay and is of great significance in realistic tracking systems. We provide details of the Gaussian mixture implementation of the proposed prior-CC-PHD filter based on a diffuse target birth model and analyze the filtering-communication parallelization. In addition, we evaluate the gain of the prior-CC scheme with respect to the filtering accuracy in comparison with the standard CC scheme via simulations using a stationary linear sensor network and a dynamic nonlinear sensor network, respectively. Published in: IEEE Sensors Journal ( Early Access ) Date of Publication: 22 June 2020 DOI: 10.1109/JSEN.2020.3004068 相关博文链接： 多源信息融合的Best Fit of Mixture原则 多目标信息融合问题 分布式网络信息共享：Many Could Be Better Than All 轻松多传感器多目标探测与跟踪！ 通讯量最小的分布式多目标跟踪器 基于多传感器观测聚类的鲁邦多传感器PHD滤波 基于受限测距传感网的分布式多目标跟踪 基于算术均值一致性的分布式伯努利滤波目标联合探测与跟踪 基于算术均值一致性的高效、分布式、联合传感定位与多目标跟踪
在几乎所有（有关传感器的）参数和（有关目标的）模型条件均未知的情况下，怎么做到：采用一个无线传感网（节点之间还可能非相互独立）去探测、跟踪未知数目的一群目标？ 传统上来说，就是采用一个传感器去跟踪估计一个目标，滤波器的设计也往往要基于准确的传感器参数（比如噪声统计特性、杂波率、漏检率等）和相对准确的目标模型信息（否则就需要构建多模型或者自适应模型进行近似或学习逼近），这些所涉及的参数和模型，任何一个未知都会给估计跟踪带来很大困难！比如常常借助于有效的系统辨识或者参数学习机制等等，滤波器才能够有效运行。 那么一堆传感器（特别是分布式网络链接起来）和一堆目标呐，什么属性都完全未知的时候呐？不仅仅是数量的升级，还可能带来传感器之间和目标之间的交互关联等复杂问题！这就使得多传感器多目标跟踪成为一个更为棘手的难题！大道至简，难到一定程度的问题也许可以用简单的方法解决！ 请看下文所提出的一中 Lazy Networking Approach: 轻松网络协作方法，只需要Flooding 和 Clustering两个操作， 就可以应对各种参数和模型未知，方法简单计算快、效果可以胜过提供了真实参数和模型信息的传统滤波器（也就是先不让滤波器去操心参数和模型未知的问题，给它们最理想的条件）！ Distributed Flooding-then-Clustering: A Lazy Networking Approach for Distributed Multiple Target Tracking Authors: Tiancheng Li ; Juan M Corchado ; Huimin Chen Abstract: We propose a straightforward but efficient networking approach to distributed multi-target tracking, which is free of ingenious target model design. We confront two challenges: One is from the lack of statistical knowledge about the target appearance/disappearance and movement, and about the sensors, e.g., the rates of clutter and misdetection; The other is from the severely limited computing and communication capability of the low-powered sensors, which may prevent them from running a full-fledged tracker/filter. To overcome these challenges, a flooding-then-clustering (FTC) approach is proposed which comprises two components: a distributed flooding scheme for iteratively sharing the measurements between sensors and a clustering-for-filtering approach for target detection and position estimation from the local aggregated measurements. We compare the FTC approach with cutting edge distributed probability hypothesis density (PHD) filters that are modeled with appropriate statistical knowledge about the target motion and the sensors. A series of simulation studies using either linear or nonlinear sensors, have been presented to verify the effectiveness of the FTC approach. Published in: https://ieeexplore.ieee.org/document/8455759 DOI: 10.23919/ICIF.2018.8455759
第三届网络与分布式计算国际会议（ ICNDC2012 ） 2012.10.21—2012.10.24, 中国 , 杭州 论文征集 全文投稿截止日期： 2012 年 5 月 1 日 论文录用通知日期： 2012 年 6 月 1 日 交修订版截止日期： 2012 年 6 月 20 日 大会主席： 黄理灿 浙江理工大学 黎建辉 中国科学院 曹军威 清华大学 郭毅可 英国帝国理工大学 刘元安 北京邮电大学 David W. Walker 英国卡迪夫大学 【会务组联系方式】 联系人： 闫志文、刘振、李雪利、刘飘悦、李志龙 电话 / 传真： + 86 - 0571-86843267 E-mail: icndc2012@inetdc.org 通讯地址： 中国浙江理工大学 邮政编码： 310018 会议网 址 ： http://www.inetdc.org/meeting/icndc2012/ 国际委员会： Mark Baker (University of Reading, UK) John Brooke (University of Manchester, UK) Rajkumar Buyya (University of Melbourne, Australia) Wentong Cai (Nanyang Technological University, Singapore) Jie Cao (Nanjing University of Information ScienceTechnology, China) Gang Chen (Chinese Academy of Science, China) Kang Chen(Tsinghua University, China) Giuseppe Ciaccio (Universita' diGenova, Italy) Philippe Cudre-Mauroux(Massachusetts Institute of Technology, USA) Jiazhu Dai (Shanghai University, China) Yong Fang (Chinese Academy of Sciences, China) Zaiwen Feng(Wuhan University, China) Haiwu He (INRIA, France) Shaoyi He(California State University at San Marcos, USA) Jinzhu Gao ( University of Pacific, USA) Weidong Geng ( Zhejiang University,China) Jinyuan Jia (TongjiUniversity, China) Keyuan Jiang (Purdue calmet University, USA) Gang Kou (University of Electronic Science and Technology of China, China ） Jianping Li (Chinese Academy of Sciences, China) Shijian Li (Zhejiang University, China) Wei Li( Chinese Academy of Sciences , China ) Xiaolin (Andy) Li (Oklahoma State University, USA) Liu Hong( Chinese Academy of Sciences , China ) Willie W. Lu （ Chairman, USCWC ， USA ） Kai Nan (Chinese Academy of Sciences, China) Daowu Pei (ZhejiangSci-TechUniversity) P.I. Poromarenko (National Mining University of Ukraine, Ukraine) Omer F. Rana (Cardiff University, UK) Zhen Shen( Chinese Academy of Sciences , China ) Yingwen Song (National Institute of Advanced Industrial Science and Technology ,Japan) Wei Tan （ Staff Researcher, IBM T. J. Watson ） Cho-Li Wang (Univ. of Hong Kong, Hong Kong) Hecheng Wang(Hangzhou Diazhi University, China) Jue Wang (Chinese Academy of Sciences, China) Xiaodong Wang (STFC, DaresburyLaboratory, UK) Yaming Wang ( Zhejiang Sci-Tech University,China) Fenghua Wen (Changsha University of Science and Technology, China) Suhong Yang (Hangzhou Diazhi University, China) Zhiwen Yu( Northwestern Polytechnical University, China) Hehua Zhang(Tsinghua university, China) Yunhua Zhang ( Zhejiang Sci-Tech University, China) Zhiming Zhao (University of Amsterdam, Netherlands) Chengxiong Zhou (Chinese Academy of Sciences, China) Ligang Zhou (City University of Hong Kong, Hong Kong) Jinlou Zhao (Harbin Engineering University, China) 会议基本信息： 网络与分布式技术是当前乃至未来的 IT 技术的重要组成部分。当下一代网络成为现实，移动通信系统将来发展到 3G 甚至 4G 时代时，将会出现传统软件向网络应用转变的趋势。为了便于工业界和学术界共同研讨网络与分布式计算的热点话题和发展趋势，我们将于 2012 年 10 月 21 日至 24 日在中国杭州举办第三届网络与分布式计算国际会议。会议重点（ 1 ）分布式计算和分布式系统方面，包括集群和网格，服务组合和业务流程，点对点对等系统，云计算等（ 2 ）网络方面，包括 IP 网络，下一代互联网，无线网络， 4G 移动通信等（ 3 ）分布式应用系统方面，包括分布式电子商务应用，分布式 e-Science 应用，管理应用系统等。 征文范围及要求： 一征文范围 包括分布式计算和布式系统，网络和分布式应用三个分支。 1 ． 分布式计算和分布式系统： 集群和网格运算，协同计算，分布式数据存储，服务组合，分布式多媒体系统，点对点对等系统，云计算，分布式资源管理，多代理系统，中间件技术，服务虚拟化，并行与分布式处理，面向服务的构架，面向服务的计算，复杂的分布式系统，面向服务的软件和系统工程，语义网络，社会网络，传感器网。 2 ． 网络：主动 / 可编程网络，移动 / 无线网络仿真，自适应网络 , Ad hoc 无线网络，分布式网络管理，无线局域网，城域网，广域网 , 3G 和智能网络，互联网络架构，多媒体网络，无线多播，传感器网络，网络隐私和安全，无线网络管理，网络服务质量和性能评价，无线协议和架构，下一代网络，新型网络架构和协议，按需网络。 3 . 分布式应用程序：业务流程整合，业务流程管理，企业资源规划，企业流程管理，协同电子商务，企业联合会，企业集成，全球企业，虚拟 / 网络企业，供应链合作，电子物流，电子商务 , B2B, B2C, C2C 模式，电子银行，电子商务，移动商务，商务数据挖掘，自适应业务，按需电子商务。 二征文要求 1. 论文须未在国内外公开发行的刊物发表。 2. 所有论文必须是英文文稿，全文不能超过五页，投稿稿件请用 Word 或 Pdf 格式排版。论文递交的文本格式：请参照 http://www.inetdc.org/meeting/icndc2012/ （ page submission ）。 3. 如果论文一经录用，所录用的论文将会被 CPS 出版，将送 EI 和 ISTP 收录引用。 最好的 50 篇论文将修改后推荐到 EI 期刊发表。 4. 所有论文文稿应提交以下电子提交系统： http://www.easychair.org/conferences/?conf=icndc2012 5. 如有任何疑问请联系 icndc2012@inetdc.org 大会主席简介： Prof. Lican Huang was graduated from Nanchang University in 1982 for B.Sc, from Hangzhou University in 1984 for M.Sc, and from Zhejiang University in 2003 for Ph.D. He has developed many large software systems in several companies, as technical leader or department manager. From 2004, he was a Senior Research Associate in Cardiff University researching at the project Workflow Optimisation for e-Science Applications(WOSE). Since 2006, he is a Director of Network Distributed Computing at Zhejiang Sci-Tech University , and guest professor at CAS and Beijing University of Posts and telecommunications. He was collected in Marquis Who’sWho in the World 2006, Marquis Who’sWho in the Science and Enginnering 2006-2007, and Marquis Who’sWho in Asia 2006-2007 due to his achievement of proposing Virtual and Dynamic Hierarchical Architecture for e-Science and Grid and VIRGO protocols. 黄理灿，男，汉族， 1962 年 1 月出生，江西乐平人，博士，教授，浙江理工大学网络与分布式计算研究所所长， International Conference on Networking and Distributed Computing(ICNDC) 网络与分布式计算国际会议主席。一直从事网络与分布式计算研究。自 2000 年开始，一直涉及 e-Science 和网格计算的研究。 发表论文 100 余篇， Sci/Ei 检索 70 余篇。担任多个项目的主持人。为多个会议主席以及多个程序委员会成员；为多个国际杂志编委。在国际上首次提出了一种新型的 P2P 网络 —— 语义 P2P 网络 --VIRGO 。 因提出 e-Science 虚拟动态分层体系结构和语义 P2P 网络协议，而被 Marquis 世界名人录（ 2006 ）、 Marquis 科学与工程名人录（ 2006-2007 ）以及 Marquis 亚洲名人录（ 2006-2007 ）收录。 曾经为多家计算机公司的技术负责人或技术经理。 2004-2006 期间在英国 Cardiff 大学担任副高级研究员（ Senior Research Associate ）。 1982 年本科毕业于江西大学（现南昌大学）； 1984 年研究生毕业于杭州大学。 2003 年于浙江大学获得博士学位。 黎建辉 ，男，博士，现为中国科学院计算机网络信息中心科学数据中心主任， 研究员，博士生导师， CODATA 中国委员会秘书长。 2007 年获得中国科学院计算技术研究所博士学位，主要从事大规模数据存储和管理和基于语义的海量数据集成等方向的研究工作，主持或者作为主要骨干中国科学院 “ 十五 ” “ 十一五 ” 信息化专项、科技部国家科技基础条件平台、科技部 863 项目、中国科学院创新方向性项目、中国科学院国际合作项目等多项。在大规模数据存储与管理、海量数据处理、数据语义集成等方面取得了一系列成果。发表论文 50 余篇，其中 SCI 收录 2 篇， EI 收录 20 多篇，在 Future Generation Computer Systems 、 Knowledge Information System 等国际期刊 3 篇， 主持完成了一项国家标准的研制，参与了 2 项国际标准的研制，申请国家专利 8 项，其中获得国家专利一项（排名第二），获得软件著作权 10 余项。 Dr. Jianhui Li,researcher,PhD supervisor,is theincumbentdirectorofScience Data Centerof Computer Network Information Center, Chinese Academy of Sciences(CAS) .He is also theSecretary-GeneralofChinaCommittee of CODATA. Dr.Jianhui Liobtained his Ph.D. degree from Institute of Computing Technology, CAS in 2007. He mainly engaged intheresearchoflarge-scaledata storage andmanagement, semantic based huge dataintegration.As a major technician,hepresided over theinformationspecial projects ofCAS during "10th Five-Year " and “11th Five-Year” plan period,the platform of thenationalscience and technology infrastructure,863projects,directivity innovation projectofCAS, internationalcooperation projects of CASand many other projects. Dr. Jianhui Li has made a series of achievements inthelarge-scaledata storage and management,huge data processing and data semantics integration and publishedmore than 50 papers,of which 2 SCI articles, more than 20 EI papers, 3international journalsinthe FutureGenerationComputer Systems, Knowledge Information System and other international journals. He also presided over the completion ofthedevelopmentofanational standardand participated inthe development of twointernational standards.He has applied 8 state patents, of which 1 has achieved national patent, and he won more than 10 software copyrights. Dr. Junwei Cao is currently a Professor and Assistant Dean of Research Institute of Information Technology, Tsinghua University, China. Before joining Tsinghua in 2006, he was a Research Scientist of Massachusetts Institute of Technology, USA. Before that he worked as a research staff member of NEC Laboratories Europe, Germany. Junwei Cao got his PhD in Computer Science from University of Warwick, UK, in 2001, where his PhD thesis was focused on Agent-based Resource Management for Grid Computing. He got his master and bachelor degrees from Tsinghua University in 1998 and 1996, respectively. Junwei Cao’s research is focused on advanced computing technology and applications. He recently participated in research projects on Cloud Computing and Internet of Things under the National 973 Basic Research Program, Ministry of Science and Technology of China. Junwei Cao has published over 100 academic papers, cited by international researchers for over 2000 times. Junwei Cao is a Senior Member of the IEEE Computer Society and a Member of the ACM and CCF. 曹军威，博士，清华大学信息技术研究院院务会副主任、研究员。美国 MIT 访问科学家（ Visiting Scientist ）。 1991-1998 ，清华大学自动化系本科、硕士毕业； 1999-2006 ，分别在英国华威（ Warwick ）大学计算机博士毕业、在位于德国波恩的 NEC 欧洲实验室任 Research Scientist 、美国 MIT 任 Research Scientist ； 2006 年回清华工作至今。 研究方向：先进计算技术及其应用：赛百平台（ Cyberinfrastructure ）、网格计算（ Grid Computing ）、分布式计算（ Distributed Computing ）、高性能计算（ High Performance Computing ）、数据管理（ Data Management ）、虚拟组织（ Virtual Organization ）管理和流程管理（ Process Management ）。 Dr. Yike Guo is a professor in computing science in the Department of Computing, Imperial College London. His research is in the areas of large scale scientific data analysis , data mining algorithms and applications, parallel algorithms and cloud computing. He graduated in Computer Science from Tsinghua University of China and has a PhD in Computational Logic and Declarative Programming at Imperial College London. During his PhD study, he was one of the founding members of the field studying uniform declarative programming by integrating functional and logic programming languages. Later, his work on functional coordination forms established a foundation for structured parallel programming. Dr. Yike Guo has been working in the area of data intensive analytical computing since 1995 when he was the Technical Director of Imperial College Parallel Computing Centre. During last 10 years, he has been leading the data mining group of the department to carry out many research projects, including some major UK e-science projects such as: Discovery Net on Grid based data analysis for scientific discovery; MESSAGE on Wireless mobile sensor network for environment monitoring; BAIR on System biology for diabetes study. He has been focusing on applying data mining technology to scientific data analysis in the fields of life science and healthcare, environment science and security. He is the Principal Investigator of the Discovery Science Platform grant from UK EPSRC and he is also the Founder and Chief Technical Officer of InforSense Limited , an Imperial College spin-out company on enterprise platform for business and scientific intelligence. 郭毅可， 1985 年本科毕业于清华大学计算机系， 1986 年硕士毕业于清华大学计算机系， 1993 年博士毕业于伦敦大学帝国理工学院计算机系。伦敦大学帝国理工学院计算机系计算机科学正教授。 刘元安 ，教授，博士生导师 , 国家 “863 计划 ” 专家、现任 北京邮电大学 电子工程学院执行院长。长期致力于下一代移动通信系统及关键技术、天线技术、异构网络接入与融合技术、电磁兼容与测量技术等领域的关键技术研究和新产品的开发工作。刘教授所领导的无线电技术与电磁兼容实验室，有高水平的科研项目，包括国家 “863 计划 ” 、国家自然科学基金、部委计划、国际合作等。主持完成各类研究课题 20 余项，在国内外学术刊物发表研究论文 100 余篇，包括 IEEE Transactions on XXXs 、科学通报、电子学报、通信学报和微波学报等，出版《宽带无线接入》等专著五本。是教育部科技委信息学部学部委员 , 中国通信标准化协会 TC-9 组副主席 , 中国电子学会和中国通信学会高级会员，电子学会电磁兼容分会委员， IEEE 会员。 Professor David W. Walker received a B.A. degree in Mathematics from Jesus College, University of Cambridge, in 1976. His M.Sc. degree in Astrophysics was obtained from Queen Mary College, University of London, in 1979, and his Ph.D. from the same institution in 1983. Professor Walker subsequently held postdoctoral appointments at the University of London and the Jet Propulsion Laboratory. In 1986 Prof. Walker became a staff scientist in the Concurrent Computation Project at the California Institute of Technology, and in August 1988 was appointed to the University of South Carolina mathematics faculty as an associate professor. In September 1990 Prof. Walker joined the Mathematical Sciences Section of Oak Ridge National Laboratory, where he headed the mathematics group, and became a senior research staff member in 1995. Since December 1995, he has been Professor of High Performance Computing in the Department of Computer Science at the University of Wales Cardiffi where he also serves as Director of the Welsh e-Science Centre. Professor Walker's research interests focus on software, algorithms, and environments for computational science on high performance computers. He has been closely involved in the development of the ScaLAPACK parallel software library, and the MPI message passing standard. He has also contributed to the design of a parallel version of the Community Climate Model, and has published a number of papers on the parallel implementation of particle-in-cell algorithms for plasma simulations. He has also been involved in the benchmarking of science and engineering applications codes on parallel computers. Professor Walker has published over 70 papers in the area of parallel computing and has co-authored three books on the subject. In 1992 he founded a series of conferences on high performance computing under the auspices of the Gordon Research Conference organization. He has organized a number of other conferences and workshops in the same area
标签: 分布式
