Stable AQM Algorithm Based on AdaptiveWavelet Neural Network Controller

Abstract

In this paper, we present the adaptive wavelet neural network (WNN) controller, which is trained by adaptive learning rates (ALRs), as an active queue management(AQM) in end-to-end TCP network. In TCP network, AQM is important to regulate the queue length and short round trip time by passing or dropping the packets at the intermediate routers. RED and PI algorithms have been used for AQM formerly. But these algorithms show weakness in the detection and control of the congestion under dynamically changing network situations. In our method, the WNN controller using ALRs is designed to overcome the problems, which adaptively controls the dropping probability of the TCP network and is trained by gradient-descent algorithm. We apply Lyapunov theorem to verify the stability of our proposed WNN controller. The simulation results show that the performance of WNN controller using ALRs is superior to that of PI controller.