Self-tuning fuzzy AQM mechanism based on traffic load and queue length
一种基于负载和队列的模糊主动队列管理算法

The existing fuzzy Active Queue Management (AQM) mechanisms mostly consider the queue length and its change rate as fuzzy inputs, while rarely consider the effect of packet arriving rate together. In order to design a more effective AQM algorithm, a novel fuzzy AQM algorithm called FQL-AQM was proposed. FQL-AQM considered both real queue length and average traffic load factor as fuzzy inputs, and regulated packet dropping probability through a fuzzy self-tuning algorithm. This algorithm can keep the queue length stable at the reference queue length level, and keep the average load factor stable at 1. And FQL-AQM can be well applied to a variety of network environments and can improve the network robustness. Simulation results show that FQL-AQM algorithm is efficient, stable and outperforms other proposed AQM algorithms in response rate, link utilization ratio, and queue length stability.