摘要： 分析现有分层组播拥塞控制的优缺点，提出一种基于模糊逻辑的分层组播拥塞控制机制FLMCC。接收端利用指数平滑预测模型预测期望速率，根据期望速率选择层数，从而获得不同的吞吐量。发送端根据接收端的反馈使用模糊控制器对发送速率进行调整，从而平滑发送速率。仿真结果表明，FLMCC具有良好的公平性和速率平滑性，能够适应网络异构性。

关键词: 分层组播, 模糊控制, 速率调节, 预测

Abstract: The merits and shortcomings of existing layered multicast congestion control mechanisms are analyzed respectively. A Fuzzy Logic-based layered Multicast Congestion Control(FLMCC) mechanism is proposed. In FLMCC, each receiver obtains different throughput by subscribing to different numbers of layers according to their expected rates, which is predicted by exponential smoothing prediction model. The sender adjusts sending rate timely to reduce the jitter of sending rate based on the feedback information from receivers by a fuzzy controller. Simulations indicate that FLMCC has fine fairness as well as smoothness, so it adapts the heterogeneous network well.

Key words: layered multicast, fuzzy control, rate adjusting, forecast

中图分类号: 

• TP393