Flow Control Based on Kalman Algorithm and Neural Network Prediction in Networks

The propagation delay in networks has a great adverse effect on rate-based flow control. Proposed in this paper is the composite control based on Kalman algorithm feedback control and neural network predictive compensation on line for ABR communication in ATM networks, which can better overcome the adverse effect caused by the delay on the control rapidity and stability. The simulation shows that the scheme can make sources respond to the changes of network status rapidly, avoid the congestion effectively, and utilize the bandwidth sufficiently. Compared with PID control, much lower cell loss rate, much higher link utilization rate, and much smaller buffer capacity can be achieved.