Periodically Adapted Burst Assembly Algorithm Based On Novel Traffic Prediction In Obs Networks

Qian Zhou,Lemin Li,Sheng Wang,Shizhong Xu,Yi Zhang
2007-01-01
Abstract:Optical Burst Switching (OBS) is under study as a promising paradigm for the next generation internet and burst assembly is one of the critical issues to be solved. A reasonable assembly algorithm should be self-adaptive to the offered load, making the aggregation dynamically optimized. However, the existing algorithms can not indeed guarantee such optimization without effective traffic prediction. Besides, their adaptation granularitics are all confined to a burst, which lacks flexibility. This paper presents a more intelligent assembly scheme termed TP-PAA (Traffic Prediction based-Periodically Adapted Assembly). It adjusts threshold periodically based on a novel traffic prediction mechanism which proves accurate and fast convergent. Simulation results validate its good adaptability and improved bandwidth utilization. Furthermore, TP-PAA enables scalable adaptation granularity to enhance flexibility.
What problem does this paper attempt to address?