Blocking probability and convergence analysis of an adaptive DQDBbandwidth balancing mechanism

摘要

The authors present a distributed realization of an adaptivenload-based bandwidth balancing mechanism (APR, adaptive proportional)nfor improving DQDB (distributed queue dual bus), fairness andnperformance. Under the APR strategy, bandwidth is assigned distributedlynand dynamically to each station according to the current offered loadndistribution in the system. The major significance of the method is itsnsimple realization in a distributed environment and its ability tonachieve load-based fairness in throughput and blocking probability whilenat the same time reducing the wasted system bandwidth to a minimum. ThenAPR scheme is shown to be fair in handling multi-priority traffic. Innaddition to the performance of APR, the system convergence behavior isnalso analyzed