Decentralized regulation for a class of large-scale networks with saturation

In this paper, we study a nonlinear large-scale network model with saturation constraints on both the control input and the state variables. We achieve buffer queue length regulation against unknown inter-node traffic interferences via a decentralized saturated control law. Motivated by physical characteristics of communication networks, two conditions regarding the inter-node traffic are discussed, namely a Lipschitz-type condition and a "PE" condition. Under these conditions, by using sliding-mode type controllers, the regulation error of every node converges asymptotically to zero for all feasible initial queue lengths. Our main (global) regulation result is based on an interesting extension of the popular Young's inequality to the case with saturation.