【摘要】 文章在分析现有提高无线TCP性能方案的基础上,提出一种新的流量控制方法,即基于显式窗口反馈的无线网络流量控制方案。在SNOOP中引入有线网络的显示窗口自适应(EWA)算法,通过对BS共享缓存的实时监测,应用模糊控制算法预测当前拥塞窗口(CWND)的大小,并显式反馈给发送端,使TCP的发送窗口能快速响应网络负荷状况的变化,避免分组的丢失。仿真结果表明,该方法增强了网络对拥塞的自适应性以及对无线信道差错的实时处理能力,提高了网络的吞吐量。更多还原

【Abstract】 Comparing to wire networks, the application of TCP over wireless networks is much different due to high bit error rates and fading and handoff. A new approach to improve transport control protocol (TCP) throughput in wireless networks is proposed, in which a fuzzy adaptive control algorithm based on BS buffer changing is adopted to calculate the congest window, and explicit feedback to TCP sender. As a result, the TCP sending window can respond to the load changes of networks rapidly, the loss of packets is avoided and the TCP flow becomes smoother. The simulation results show that this method can strengthen the adaptation and channel errors processing capability in real time and improve the throughput of the network, and this control scheme is more robust and adaptive than normal Snoop.更多还原