RED的网络性能研究

Network Performance of RED

在很多网络中,支持新的拥塞避免机制(如RED和ECN)的节点将会和支持传统丢尾缓存 (Drop Tail)的节点共存。研究含有两个拥塞节点的混杂网络中TCP流集的性能,比较了几种控制算法的数据流公平性,仿真结果显示采用Adaptive算法并不能增加数据流的公平性;仿真也显示混杂网络中单个RED参数的设置情况会影响整个网络的性能,具体可以从网络吞吐量、响应时间等判断。RED的参数设置在双瓶颈情况下再次变得重要,正如在单瓶颈的情况下一样。

In many networks, the nodes supporting new congestion avoidance mechanisms (such as RED and ECN) will exist with the nodes supporting traditional Drop-Tail mechanism. This paper studies the performance of the TCP flows in the heterogeneous network including two congested nodes, and compares the flow fairness of some control algorithms. The simulation results show that the parameters’ setting of single RED router will affect the performance of the whole network, and this can be judged from the network throughput, response time, etc. The parameter setting of RED in double bottleneck network becomes as important again as in the single bottleneck network. With the fixed parameters, RED is difficult to provide the performance better than Drop-Tail.