Fixed-time Congestion Tracking Control for a Class of Uncertain TCP/AQM Computer and Communication Networks

摘要

Abstract The fixed-time congestion control problem of TCP/AQM networks with UDP flows and un-modeled uncertainties has been studied thoroughly in this paper. The fluid-flow based TCP/AQM network system model is analyzed, and a fixed-time adaptive congestion control algorithm is proposed for the nonlinear network system model with unknown UDP flows and un-modeled uncertainty. The new control algorithm was derived using combined requirements for the fixed-time control and the prescribed performance control by employing the back-stepping technique and the fuzzy approximation methodology. The universal approximation property of fuzzy-logic system models is used to deal with the external disturbance and the uncertainty that may appear in the network system. Using the prescribed performance technology, the transient and steady state performances of the tracking error are shown to have been enforced to meet the design requirements. Using the back-stepping methodology, a fixed-time AQM controller is designed so as to ensure that the fixed time of the closed-loop system response remains bounded. The obtained simulation results demonstrate the extent of effectiveness and superiority of the proposed design relative to known previous results.