On Active Queue Management Algorithms For TCP Networks Based On Event-Triggering Scheme

With the increase of network scale and the continuous growth of network users and applications,the network congestion has become an important problem.The router-based active queue management(AQM)mechanism,which combines with port-based congestion control mechanism,can effectively solve the problem of congestion control for the TCP networks.At present,the research on AQM algorithm based on control theory is an active issue.In order to reduce the unnecessary waste of the limited network bandwidth resources,the thesis presents some AQM algorithms based on event-triggering.The main research conclusions are as follows:Firstly,for the linear TCP network systems with uncertainties and external disturbance,the co-design of event-triggering controller and event-triggering condition is given.Besides,the co-design of robust event-triggering controller and event-triggering condition is also given in the thesis.The gain of the designed controller and event-triggering parameter can be calculated by linear matrix inequalities(LMI)technology.Simulation results illustrate the effectiveness of the designed controllers.Meanwhile,the algorithm can effectively reduce the trigger frequency of controllers,so it can reduce the network resources.Secondly,an observer-based AQM algorithm is proposed because the system state is not fully measurable.Considering the uncertainty and external disturbance of network,the robust event-triggering controller and event-triggering condition based on observer are designed and two methods to calculate the gain of the designed controller are given.The simulation results show that the algorithm is strongly robust and it can reduce the calculating frequency of controller.Thirdly,because the T-S fuzzy model can suitably represent a nonlinear system,the T-S fuzzy controller based on event-triggering and event-triggering condition are designed.The effect of uncertainties and time delays to networks are considered in the algorithm.By using LMI technology,the sufficient condition which can ensure the stability of network is given.The simulation results show that the algorithm can ensure the stability of network system and it can also reduce the network resources.Finally,the summary of the thesis is given and the research directions in future are also proposed.