Congestion Control Algorithm Based On Ts Fuzzy Control Theory Network

With the rapid development of the computer network,soaring increase of the Internet users and constant appearance of the new network applications,have brought a series of problems to the network's normal operation,one of the most serious problem is network congestion.When the network congestion occurs,it will lead to the network throughput's decrease,the delay's increase and the packet loss's increase, seriously even cause network collapse.Therefore,in order to prevent and control the network congestion,we need to design the effective network congestion control algorithm. Currently,the network congestion control algorithm is the important means for improving the network's performance, ensuring the network's stable operation and enhancing the network's QoS(Quality of service).In recent years,experts and scholars have achieved lots of achievement in the field of network congestion control algorithm's research.Due to the end-to-end TCP congestion control has certain limitation,Active Queue Management(AQM)algorithm has become the current research hotspot in the network congestion control's field.This article mainly researches the network congestion from the aspect of AQM algorithm. The main research results are as follows:(1) An AQM algorithm is proposed based on T-S(Takagi-Sugeno)fuzzy model. For the nonlinear model of the network congestion control system, the T-S fuzzy model of the congestion control system is established for using T-S fuzzy model's feature which can approximate nonlinear systems very well,and the state feedback AQM control algorithm is designed, the conditions of the system's stability are given,and the design method of the AQM controller's parameters is given by using LMI(Linear Matrix Inequality) techology. Simulation results show that this algorithm can maintain the queue length around the given target,has good stability and robustness. Comparison with other AQM schemes, such as PI,REM and ARED,has demostrated the superiority of this algorithm in convergence rate, stability, robustness and the fluctuation of queue length.(2) A single neuron adaptive PID control algorithm is designed based on T-S fuzzy model.Through introducing the T-S fuzzy model to adjust the neuron's gain,the single neuron adaptive PID control algorithm has the fuction that the gain can auto-tune. Simulation results show that,comparison with the improving single neuron adaptive PID control algorithm,this algorithm has better convergence,can achieve faster convergence to target queue length,and shows a better stability, robustness and smaller steady-state error.