N次随机丢包的被动队列管理算法  

Passive Queue Management Algorithm Based on Random Dropping N Times

作　　者：王文涛[1] 郭峰[1] 王奇枫[1] 郑芳[1] 唐菀[1] 

机构地区：[1]中南民族大学计算机科学学院,武汉430074

出　　处：《中南民族大学学报（自然科学版）》2014年第2期100-105,共6页Journal of South-Central University for Nationalities：Natural Science Edition

基　　金：国家自然科学基金资助项目(61103248)

摘　　要：针对主动队列管理参数设置敏感、消耗大量资源等问题,提出了N次随机丢包的被动队列管理算法.当队列满时,根据当前网络中与路由节点链接的发送端数目,决定丢弃数据包N的值,然后再采用改进的随机丢弃策略丢弃N个数据包.新的算法与其他被动随机丢弃策略相比,克服了弃尾队列管理的缺陷,能够更快地响应网络拥塞,提高了网络传输性能,同时也能够在一定程度上保证公平性,NS2仿真表明了算法的有效性.Active queue management is sensitive to parameter settings and many resources consumption. Therefore, a passive queue management algorithm based on random dropping N packets is put forward. The number of dropped packets are decided by the number of sending nodes connected to router when the buffer is full. Then this new algorithm randomly drops the corresponding N packets with the improved random dropping strategy, it overcomes the shortcoming of drop-tail queue management and reacts faster to the network congestion than other random dropping algorithm. At the same time, the network transmission performance is increased which also ensure the network fairness to a certain extent. The simulation based on NS2 proves it＇s efficiency.

关 键 词：网络拥塞 被动队列管理 N次随机丢包 NS2仿真 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]