Based On The Intelligent And Robust Control Of Active Queue Management Algorithm Research

The rapid development of network information age is prompting the growth of variety of applications. The congestion phenomenon seriously affect network quality of service and resource utilization. Because the source-side control mechanism can not meet the needs of the current user, based router’s link-based algorithm join in network congestion control method. AQM algorithm is the most effective mechanism to deal with congestion in the link algorithm. To resolve the sensitive parameter settings, the model does not match and the anti-disturbance capacity,the paper use intelligent method and robust control theory and put forward the following three algorithms:(1) The past neuron PID-based AQM mechanism ignores the impact of the proportional gain values. However, the proportional gain is an important factor affecting the system response capabilities. Therefore, this paper presents adaptive AQM algorithm optimization based on improved single neuron.This chapter use supervised Hebb algorithm to adjust the weighting coefficients and then use the steepest descent method tuning neuron PID proportional gain. This paper also gives a stability analysis of the algorithm. Simulation results show that the algorithm has a faster response speed.(2) For dynamically changing network congestion, this paper presents based the parameter uncertainties and Delay AQM algorithm. The chapter focus on parameter uncertainty, the state delay and other factors in the model. Thougth select a suitable Lyapunov function, we get independent of the delay control law in order to guarantee the stability of the system. Finally, the simulation results show that the queue of the algorithm can stabilize in the near ideal value.(3) In order to improve the anti-jamming features of AQM controller, this paper presents discrete AQM algorithm. The present of robust AQM controller can not eliminate the steady state error. The paper add design integral part in the design of AQM controller to greatly improve the control performance. We focus on variable delay of the state and the input, the uncertainty of the network parameters,external disturbance.Finally we get the delay dependent stability condition. In the changing network environmentm, the algorithmt has superior stability.