大时滞网络自适应预测PI主动队列管理算法

Adaptive Predictive PI Algorithm for Active Queue Management in Large Delay Network

针对网络中存在的大时滞和网络参数时变问题,提出一种自适应预测P I主动队列管理算法.将Sm ith预估器与达林算法相结合,既克服了大时滞带来的不利影响,也减少了控制器参数整定数量.利用网络参数与控制参数所具有的确定关系,通过在线估计网络参数来实时调节控制参数,使得控制器能够适应网络参数的变化,同时采用线性化方法分析了系统局部稳定性.仿真结果表明,所提出的算法是可行而有效的.

An adaptive predictive proportional integral （APPI） algorithm for active queue management （AOM） is proposed to cope with the large delays and time-varying network parameters. Smith predictor and Dahlin algorithm are combined to alleviate the unfavorable influences of time-delay and reduce the number of the tuning parameters. The control parameters are automatically tuned according to on-line estimation of link capacity and traffic load, which makes APPI perform well for a wide-range of network conditions. The local stability of the system is analyzed using the linearization method. Simulation results show the effectiveness of the APPI algorithm.