摘要
提出了一种新的基于网络丢包率的动态自适应的主动队列管理的改进算法。该算法首先用早期网络的丢包率标记到达的数据包,并作为数据包的丢弃概率。这样使得到达数据包的丢弃概率逼近当前实际网络的丢包率,然后经过自适应调整丢弃概率使得缓冲队列长度保持在一定范围内。通过NS仿真实验表明了该算法可以很好的控制队列长度,降低延迟抖动。
A new dynamic adaptive AQM（Active Queue Manngement）scheme based on network loss ratio is proposed. The packet loss ratio of the early network is adopted to mark the arriving packet is marked, which is used as the early drop probability so that the probability can approach the loss ratio of current real network. By adjusting the drop prob- ability adaptively, the length of buffer queues can be kept in a certain area. The simulation experiment results using NS software tool show that the scheme can flexibly control the length of queue and reduce the jitter of delay.
出处
《微电子学与计算机》
CSCD
北大核心
2007年第2期24-27,共4页
Microelectronics & Computer
基金
国家自然科学基金项目(60572074)
关键词
队列管理
丢包率
拥塞控制
拥塞避免
Queue management
Packet loss ratio
Congestion control
Congestion avoidance
