有线/无线网络中最小方差拥塞控制算法

Congestion control algorithm based on minimum variance theory in wired/wireless network

为了满足流媒体等新型应用对网络服务质量的需求,该文提出了基于最小方差理论的网络拥塞控制算法。针对有线/无线网络特性,利用平衡点线性化方法将拥塞控制模型转换为离散传递函数,并考虑系统噪声和随机性的影响,将广义最小方差控制理论引入有线/无线网络的拥塞控制中,设计了基于最小方差的网络拥塞控制算法。仿真实验表明,在变化网络环境下,所设计的网络拥塞控制算法是有效和鲁棒的,能够维持期望的队列长度,有效抑制队列长度波动和延时的影响,提高有线/无线网络服务质量。

In order to meet the demands of the novel application such as the streaming media for the network service quality, the network congestion control algorithm based on the minimum variance theory is proposed here. According to the characteristics of the wired /wireless network,the congestion control model is converted into the discrete transfer function near the equilibrium point. In view of the system noise and random effects, the congestion control method based on the generalized minimum variance theory is designed. Simulation results indicate that the proposed algorithm is effective and robust. In the variable network condition, the algorithm can maintain the reference queue length and restrain the jitter of the queue length and the delay, and can improve the quality of service in the wired/wireless network.