Periodically Adapted Burst Assembly Algorithm Based on Novel Traffic Prediction in OBS Networks

摘要：

Optical Burst Switching (OBS) is under study as apromising paradigm for the next generation internet and burstassembly is one of the critical issues to be solved. A reasonableassembly algorithm should be self-adaptive to the offered load,making the aggregation dynamically optimized. However, theexisting algorithms can not indeed guarantee such optimizationwithout effective traffic prediction. Besides, their adaptationgranularities are all confined to a burst, which lacks flexibility.This paper presents a more intelligent assembly scheme termedTP-PAA (Traffic Prediction based-Periodically AdaptedAssembly). It adjusts threshold periodically based on a noveltraffic prediction mechanism which proves accurate and fastconvergent. Simulation results validate its good adaptabilityand improved bandwidth utilization. Furthermore, TP-PAAenables scalable adaptation granularity to enhance flexibility.

作者: Qian Zhou Lemin Li Sheng Wang Shizhong Xu Yi Zhang

作者单位: School of Communication and Information Engineering University of Electronic Science and Technology ATD, Wireline Networking BU Huawei Technologies Co.Ltd.Shenzhen, CHINA

会议类型: 国际会议

会议名称: 2007年通信、电路与系统国际会议(2007 International Conference on Communications,Circuits and Systems Proceedings)

会议地点: 日本福冈

会议语种:英文

在线出版日期: 2007-07-11（万方平台首次上网日期，不代表论文的发表时间）