一种基于模糊理论的拥塞控制算法

路由器在处理拥塞时,通常把网络载荷分为正常、拥塞避免和拥塞三个阶段,并在拥塞避免阶段开始采取行动。由于各个阶段的描述存在一定的不确定性,因此一些传统的方法很难取得更好的效果。论文则充分利用模糊理论在处理不确定性问题上的优越性,采取整体和局部相结合的方法,从而在保证各TCP连接吞吐量的公平性方面,取得了比传统的方法更好的效果。实验仿真结果表明,模糊算法较传统方法更为合理,从而更好地改进了路由器的拥塞控制性能。

A Fuzzy Congestion Control Algorithm in Routers

With the development of Internet applications,it is increasingly important for router itself to improve the abil-ity to deal with network congestion.Usually in the process of congestion control there are several phases including nor-mal,congestion avoidance and congestion.Because of the uncertainty in distinguishing the three phases from each other,some traditional methods are not able to obtain satisfactory effect.However,fuzzy congestion control algorithm based on fuzzy theory utilizes fully the predominance of fuzzy theory in dealing with the uncertain event ,then achieves fair throughput by combining entireness with part.Simulation results show that the proposed algorithm can obtain better per-formance compared with traditional schemes.