自适应PD主动队列管理算法

An active queue management algorithm based on adaptive PD controller

网络拥塞控制存在参数时变和非线性等问题,固定参数配置的PID很难达到令人满意的控制效果,针对这个不足,提出1种自适应PD(adaptive proportional and differential,APD)主动队列管理算法。该算法根据队列误差以指数形式动态调节控制器参数来改善控制性能。仿真结果表明:与PI,RPI和FREM算法相比,该算法收敛速度较快,鲁棒性较强,且队列抖动较小。

Both parameter time-varying and nonlinear problems exist in network congestion control.PID with fixed parameters has been difficult to achieve satisfactory control performance.To solve this shortage,a new active queue management scheme with adaptive PD(adaptive proportional and differential,APD) was proposed.Based on queue error,the algorithm exponentially adjusts controller parameters to improve control performance.Simulations show that compared with the PI,RPI and FREM algorithms,the proposed algorithm has faster convergence rate,stronger robustness and less jitter.