时延通信网络的新型指数稳定跟踪控制器设计

针对一个具有单瓶颈和多用户的高速通信网络,提出了一个新的指数稳定跟踪控制器的设计方法。将闭环数据率控制系统描述为一个连续时间时变时滞系统,基于线性矩阵不等式(LMI)技术和Lyapunov方法给出了闭环系统时滞依赖的指数稳定充分条件。导出了状态反馈控制器的设计方法,使得实际网络带宽分配率和实际队列长度在该控制器的作用下分别以指数速度跟踪上给定的参考值,从而有效降低了数据包丢失率,提高了网络利用率。通过在闭环系统指数稳定条件的导出过程中引入自由变量的方法使得得到的时滞依赖条件具有更小的保守性,并且闭环系统具有更快的指数稳定跟踪速度。仿真算例验证了本文方法的有效性和优越性.

New Exponential Tracking Stabilization Controller Design for Communication Networks with Time Delays

A new exponential tracking stabilization controller design method was proposed for the high-speed communication networks with one bottleneck node and multiple sources.The closed-loop data rate control system was modeled as a continuous time systems with time-varying delays,and sufficient delay-dependent exponential stability condition was derived based on the LMI technique and the Lyapunov method.State feedback controller design method was proposed,and the network bandwidth allocation fairness and the queue length could track respectively to the corresponding references with a certain exponential rate under the effect of the designed controllers.Delay-dependent criterion of less conservatism and controllers with faster tracking speed were derived by introducing free variables in the derivation the exponential stability condition.Illustrative example demonstrates the effectiveness and the superiority of the proposed method.