一种基于模糊逻辑的主动队列管理算法

A Fuzzy Logic Approach for Active Queue Management

主动队列管理 (ActiveQueueManagement,AQM)技术作为Internet拥塞控制的一种有效方法 ,对于提高In ternet的服务质量具有十分重要的作用 .本文根据TCP拥塞控制算法基于数据包丢失的窗口变化机制 ,设计了一种基于模糊逻辑的主动队列管理算法 .该算法依据路由器中队列长度的变化情况 ,根据一定的模糊自校正原则来调整数据包的丢弃概率 ,从而使路由器中的队列长度稳定在参考值附近 .仿真结果表明该算法不但十分有效 ,而且对不同的网络状况具有很好的适应能力 .

As an effective method for congestion control,Active Queue Management plays an important role in improving the QoS of the Internet.An effective self-tuning fuzzy queue management algorithm based on the mathematical model of TCP congestion control mechanism was presented.With the application of this algorithm,routers in IP network regulate its packet drop probability through a fuzzy self-tuning algorithm according to the queue length in the buffer.The main advantage of this algorithm is that the queue length can keep stable in a variety of network environments without the difficulty of parameter configuration.Simulations show that this fuzzy queue management algorithm is efficient,stable and outperforms other proposed queue management algorithms.