摘要
针对高速通信网络的拥塞问题,设计了智能PID控制器对ATM网络中的ABR速率进行实时、在线调节.引入了Smith预估补偿器来消除往返传播时延对控制造成的不利影响.采用改进的粒子群优化方法对PID控制器的参数进行调节,所设计的控制算法不依赖于网络的数学模型,克服了ABR带宽、用户数量及传播时延等不确定因素对网络性能造成的影响,并确保了带宽的公平分配及链路的充分利用.仿真结果表明,闭环系统具有良好的稳态性能及抗干扰能力,网络的服务质量得到了保证.
An intelligent PID controller is designed to adjust ABR traffic rate on-line in ATM network for the congestion of high-speed communication networks with a Smith predictor introduced in to avoid the adverse effect of roundtrip propagation delay on control. The improved PSO algorithm is introduced in tuning the parameters of PID controller. The controller designed is independent of the mathematic model of networks, thus getting rid of the adverse effects of such uncertainties as the number of users, ABR bandwidth and propagation delays on network performance and ensuring the fair distribution of bandwidth and full utilization of links. Simulation results show that the control system is stable and robust with QoS guaranteed.
出处
《东北大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2007年第1期14-18,共5页
Journal of Northeastern University(Natural Science)
基金
国家自然科学基金资助项目(62074009)
教育部暨辽宁省流程工业综合自动化重点实验室开放课题