主动式队列管理技术是IETF为了解决Internet拥塞控制问题而提出的一种路由器缓存管理技术。本文使用了目前应用较为广泛的网络仿真器NS-2,对AQM算法RED和BLUE的性能在基于NS-2仿真实验的基础上进行了比较研究,研究的性能包括在突发流情况下平均队列长度、丢包率和吞吐量等,并在此基础上对BLUE算法在突发流情况下队列稳定性进行改进。仿真结果表明在应对突发流方面,BLUE算法的性能优于RED算法,改进的BLUE算法能够更好地保持队列的稳定性。
To address the problems of TCP end-to-end congestion control, IETF proposes deploying active queue management mechanisms in the TCP/IP networks. This paper uses the widely used network simulator NS-2, and analyzes the performance of two classical AQM algorithms RED and BLUE based on the simulation data. The performance metrics include average queue size, packet loss rate and throughput. On this basis, the pa-per proposes an improved BLUE algorithm. The result shows that in the sudden flow, BLUE algorithm has a bet-ter performance than RED algorithm and the improved BLUE algorithm can keep the queue’s stabilization bet-ter than BLUE algorithm.