提出了一种动态概率优先级算法DPP,针对一类对延时和丢包率要求相对较高的应用,根据AF1队列长度动态调整概率计算参数p,有效地解决了由于突发流量带来的QoS性能下降问题。不同实验环境下的仿真结果表明,DPP算法有效改善了突发性对分组平均排队延时的影响,减少了分组丢包率。
This paper proposed a dynamic probabilistic priority scheduling algorithm-DPP. DPP algorithm assigned the parameter p used to calculate scheduling probability according to the length of queue for AF1 PHB, which could effectively solve the problem of QoS degenerateness caused by the burst traffic. The simulation results in different environment indicate that DPP effectively diminishe the effect on average queueing delay caused by the burst and reduce the packet loss ratio.