论文部分内容阅读
针对随机早期检测(RED,Random Early Detection)算法存在的公平性问题,提出了一种基于负载的公平性主动队列管理算法(LFED)。该算法通过引入网络负载和队列这两个概念,有效地判断当前网络拥塞程度,同时使用改进的丢包率公式,并且借鉴CHOKe算法的惩罚机制对非响应流进行有效惩罚,以保证不同数据流之间的公平。仿真实验表明,与RED和CHOKe相比,LFED可以得到稳定的瞬时队列长度和减少分组丢弃率,具有较好的公平性和稳定性。
Aiming at the fairness problem of Random Early Detection (RED) algorithm, a load-based Fairness Active Queue Management Algorithm (LFED) is proposed. By introducing two concepts of network load and queue, this algorithm can effectively judge the current network congestion, and at the same time, use the improved packet loss rate formula, and learn from CHOKe algorithm penalty mechanism to effectively punish non-response flow to ensure that different data streams Between the fair. Simulation results show that, compared with RED and CHOKe, LFED can achieve stable transient queue length and reduce packet discard rate, which has good fairness and stability.