摘要
针对无线网络应用,能够有效地使多路的TCP数据业务均衡地使用网络共享资源,需要系统能根据无线网络的容量实时变化,动态调整各路TCP协议拥塞窗口大小,其参数结构复杂,难于建立精确的数学模型。为此提出了一种基于BP神经网络的控制算法,利用它可有效解决多路无线TCP数据业务拥塞窗口的自适应控制问题,为基于无线网络数据业务应用的优化提出一种新的思路。
For the wireless application environment the system would be adjusted refering to the real - time capacity of wireless network, so as to make multiple wireless TCP applications to share the network resource effectively. However it was too complex to establish the exact mathematic model with all multiple wireless TCP links parameters. This paper designed one balanced- control algorithm based on neural network to resolve the adaptive control problem of muhichannel TCP data service congestion window size, which suggested one optimization method for the wireless network application.
出处
《无线电工程》
2006年第11期1-4,共4页
Radio Engineering
