Adaptive sliding mode control for TCP networks with input saturation

For the congestion control of the transmission control protocol (TCP) networks in presence of input saturation and the lumped uncertainty, an active queue management algorithm is proposed based on adaptive sliding mode control. One adaptive law is designed to estimate the lumped uncertainty such that the need for knowing the bound in advance is eliminated. The other adaptive law is proposed to ensure the system stability considering the TCP networks subjected to input saturation. Using these adaptive laws, an adaptive sliding mode controller is presented to guarantee asymptotic convergence and robustness against the lumped uncertainty. Numerical simulations are performed to show that the proposed scheme is effective in various network scenarios.