On TCP/AQM Algorithms Based On Global Sliding Mode Control

With the advent of the era of big data and the era of media,the continuous growth of users and applications on the network has brought tremendous pressure on the network,which makes network congestion an inevitable problem.The combination of router-based active queue management mechanism and TCP protocol congestion control mechanism is the main way to solve network congestion problems.Therefore,the research on the AQM algorithm based on the TCP network model is a hot issue in the current TCP network field.The TCP network can be regarded as a nonlinear system.It is a feasible and efficient method to study the congestion control by the control theory.In this paper,based on the fluid flow theory of TCP nonlinear dynamic network model,the active queue management algorithm is studied by the method of control theory.The main research work is as follows:Firstly,in view of the frequent occurrence of TCP network congestion,considering the existence of uncertain parameters and non-TCP flow interference in TCP network,an active queue management algorithm based on adaptive global fuzzy sliding mode control is designed for TCP nonlinear dynamic network system which is using GAIMD strategy.Global sliding mode control can eliminate the arrival phase in traditional sliding mode control,so that the whole response process of the system has insensitivity to parameter uncertainty and external interference.Considering that the system uncertainty term model is difficult to determine,the fuzzy approximation method is used to approximate the uncertain term,and the adaptive law is designed to adjust the approximation error online.The designed controller can not only ensure the robustness of the whole process of the sy stem,but also weaken the chattering of the sliding mode control.Simulation result shows that the algorithm has good stability and robustness.Secondly,in order solve the finite-time congestion control problem of TCP network system,considering the uncertainties and external disturbance factors in TCP network system,the active queue management algorithm based on adaptive global terminal sliding mode control is designed for nonlinear dynamic TCP network model with GAIMD strategy.The terminal sliding mode can ensure that the system converges to the origin within a finite time,and the global sliding mode can ensure the robustness of the whole process of the system.The combination of the terminal sliding mode and the global terminal sliding mode can ensure the global robustness and guarantee the limited time convergence.Considering that the upper bound of the system uncertainty is difficult to obtain,the adaptive law is designed to adaptively approximate the upper bound of the system uncertainty.Applying the upper bound of the adaptive approximation to the control law,which can eliminate the bad influence caused by system uncertainty.The simulation results show that the algorithm has good overall process robustness and fast transient response.Finally,the main work of this thesis is summarized,and some future research directions are put forward.