一种自适应主动队列管理算法ABlue

摘要

Blue是一种典型的主动管理算法,但其队列和延时波动较大,尤其在连接数较大或连接数突然发生变化时容易造成队列溢出或空闲.为此,提出一种自适应主动队列管理算法--ABlue,利用平均队列长度预测拥塞,根据网络流量的变化动态计算标记概率,从而保持队列稳定、减小延时抖动并降低丢包率.仿真实验结果表明,相比Blue算法,ABlue算法具有较低的丢包率和较高的带宽利用率.%Blue is a typical active management algorithm, but it suffers from queue size oscillations and delay oscillations, particularly the number of connections is so large or number of connections changes so suddenly that causes queue overflow or empty.Aiming at this problem, this paper proposes an adaptive active queue management algorithm——ABlue(Adaptive Blue), in order to stabilize the queue length and delay jitter, and reduce the packet loss rate according to the dynamic calculation of network flow changes mark probability.Simulation results show that ABlue algorithm significantly reduces the packet loss rate of Blue algorithm, and performs slightly better than Blue in terms of link utilization.