Adaptive Robust Control for a Class of Nonlinear TCP Dynamic Network Systems

This paper presents an adaptive robust control scheme for a class of nonlinear TCP dynamic network systems. By introducing the interference, time-delay and uncertain parameters to improve the nonlinear model, the objective of the work described here was to attenuate any various interference, time-delay and uncertain parameters factors in the actual network running environment. Using proper coordinate transformation, evaluation signal, energy storage function and adjusting appropriate weights, the systems satisfy the nonlinear L₂ gain performance index. The robust controller achieves the desired queue size and guarantee asymptotic stability. This way ameliorates the instability phenomena, when the parameters of system were changed and subjected to the external interference. Finally, the simulation results confirm the feasibility and effectiveness of the designed scheme.