一种基于模糊控制的参数自适应RED改进算法  

An improved parameter adaptive RED algorithm based on fuzzy control

作　　者：任金霞 温春晖 王水泉 

机构地区：[1]江西理工大学电气工程与自动化学院,江西赣州341000

出　　处：《微型机与应用》2017年第18期77-79,83,共4页Microcomputer & Its Applications

基　　金：江西省自然科学基金项目(20151BAB207041)

摘　　要：考虑到传统随机早期检测(Random Early Detection,RED)算法在较强的业务突发度和较大流量抖动的情况下很难获得令人满意的吞吐量这一问题,基于模糊控制理论设计了一个模糊控制器,以提高系统在减少队列长度、降低丢包率中的作用。同时由于在网络拥塞控制中传统RED算法存在着参数敏感、稳定性差等问题,故在系统中加入一个参数自适应算法,用来稳定队列长度。仿真结果表明该算法在减少队列长度、降低丢包率、提高鲁棒性方面的优化有着明显的效果。Considering the problem that the traditional Random Early Detection（ RED） algorithm in the stronger business under the condition of sudden and large flow jitter is difficult to obtain satisfying throughput,this article designs a fuzzy controller to improve the system performance in reduce the queue length and the reduce packet loss rate based on fuzzy control theory. At the same time,due to network congestion control in traditional RED algorithm has sensitive parameters baol instability,this paper introduces a parameter adaptive algorithm to the system to stabilize the queue length. The simulation results show that the algorithm has obvious effect in reducing queue length,reducing the packet loss rate,and improving the robustness of optimization.

关 键 词：拥塞控制 模糊控制 RED算法 自适应 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]