摘要
主动队列管理算法的稳定性是实现拥塞控制的基础,针对目前PID-AQM控制器控制参数大多基于经验、缺乏稳定性分析的问题,提出了粒子群优化PID参数的整定策略,同时运用稳定分析论分析PSO-PID队列管理算法的稳定性.考虑TCP/AQM系统的时滞性,将时延系统利用泰勒展开转化为非时延系统进行稳定性判定.通过NS2仿真,结果表明,PSO-PID算法稳定性明显优于PID算法.
The stability of the active queue management is the basis of the congestion control, most of the control parameters of the PID-AQM controller based on experience, lack of stability analysis. In order to solve the problem, particle swarm optimize the PID parameters strategy was proposed, while use the theory of stability to analysis the stability of the PSO-PID active queue management. Consider the TCP/AQM system delay, the paper transfers the delay system into non-delay system by using Taylor series expansion. The NS2 simulation results the PSO-PID algorithm shows some advantages over the PID algorithm.
出处
《微电子学与计算机》
CSCD
北大核心
2014年第2期106-109,共4页
Microelectronics & Computer
基金
国家自然科学基金项目(60974129)
关键词
拥塞控制
主动队列管理
粒子群优化
PSO-PID控制器
稳定性
congestion control
active queue management
particle swarm optimization
PSO-PI controller
stability domain
