Adaptive Generalized Minimum Variance Control for TCP/AQM

A new controller AQM strategy for dynamically varying TCP/AQM networks is proposed. Adaptive generalized minimum variance is Combination of on line parameters estimations and generalized minimum variance (GMV). The GMV is a suitable control method especially in the presence of measurement noise, but when the process dynamics changes, the controller performance declines. To overcome this problem, the adaptive GMV control approach is recommended. Controller performance was evaluated in two modes, TCP network with unknown parameters and TCP network with time varying parameters. Simulation results indicate that the adaptive GMV AQM (in either case) has been able to keep queue length around the desired point.