报告时间:2015.6.1(星期一)下午2:00
报告地点:电子楼107
报告人:Prof Sanjay Kumar Bose
邀请人:沈纲祥 特聘教授
报告摘要: Rateless Codes are effective in reducing the costs of retransmissions over erasure channels. Some rateless codes have systematic leading parts which have been shown in the literature to enhance their performance. The systematic part allows a node to start transmitting packets as soon as they are received without waiting for the entire message to be decoded. We show how simple analytical models can be constructed to estimate link costs for such codes. Network routing protocols for systems where links use rateless codes for unicast transmissions are presented and their performance is discussed. It is also possible to reduce the end-to-end delay for unicast applications using special versions of these protocols for streaming applications. Using systematic rateless codes, we propose a network routing algorithm to reduce streaming delays. Its performance is studied and compared against existing algorithms through simulations in network scenarios where the nodes use systematic rateless codes for communication. We also present some initial research results on approaches that can be followed for network routing for multicast applications in such systems.
报告人简介:Professor Sanjay Kumar Bose received his bachelor degree from the University of Indian Institute of Technology, Kanpur, India, M.S. and Ph.D. degrees from Stony Brook University, USA. He is currently the professor, Department of EEE, I.I.T. Guwahati. He was the associate professor, School of EEE, NTU, Singapore, and the Senior Fellow, Nanyang Technological University, Singapore.
Professor Sanjay Kumar Bose is a Senior Member, IEEE (Chairperson, IEEE Uttar Pradesh Section, 1996, 1997), Member of Sigma Xi, Member of Eta Kappa Nu, Fellow of IETE (India). His research interests include Modeling, Simulation and Analysis of Communication Networks, Performance of WDM Systems, ATM Networks, Queuing Networks, Network Applications, Multiple Access Schemes, Mobile Satellite Systems, Communication & Queueing Related Software Development.