Research On Congestion Control And Dynamics Analysis Based On Network Structure Characteristics

Due to the rapid growth and the unbalanced distribution of cyber source and load,Internet's"best effort"service mode determines that the network congestion is inevitable.It will not be eliminated because of improving the processing ability and increasing the network bandwidth.Therefore,the effective congestion control mechanism is very critical to guarantee the QoS?Quality of Service?.As a result,in the past twenty years,network congestion control is always an important and challenging problem that has attracted continuous attention from a lot of researcher around the world.Although great deals of valuable achievements have been obtained,but there is still much work need to be further improved.From the point of view of network topology,this paper is focused on the congestion control modeling of single routing and multi routing network topology,congestion control algorithm design and the inherent dynamic analysis and control of TCP/AQM congestion control system in wireless access network.The main work and innovation mainly include the following aspects:?1?Aiming at the problem that the model error is existed when the nonlinear model of congestion control system is linearized,the system modeling based on T-S fuzzy theory is studied.And the congestion control algorithm is designed based on the new model.Firstly,the T-S fuzzy model is proposed to describe the wired single route TCP/AQM network.Considering the asynchronous problem of the antecedent variable,the relationship between the controller and the antecedent variable is built.Using the Lyapunov-Krasovskii function and cone complementarity linearization method,the fuzzy state feedback controller that can ensure that the system is asymptotically stable in the presence of parameter variations and UDP data streams is designed.Secondly,the research object is expanded from the cable network to the wireless access network.The change in link is seen as the system disturbance,and the T-S fuzzy model of the uncertain delay wireless access network congestion control system is established.Based on the model,the robust H_? controller is designed to ensure that the system is robust asymptotically stable and satisfies certain performance index.?2?Aiming at the problem that the existing congestion control algorithms refer to multi routing environments mostly consider single node congestion control,ignore the communication between nodes,and lack the overall research.Based on directed graph theory,the network congestion control model describing the environment of multi router is established.The congestion control algorithm is designed using consistency theory.Firstly,by analyzing the dynamic characteristics of network nodes?data sending end,router and data receiving end?and link,a network congestion control model in multi router environment is established based on directed graph theory.Secondly,through analyzing the nature of network congestion,the network congestion control is transformed into the consistency problem of a data transmission rate.The congestion control algorithm is designed.Simulation results show that the proposed congestion control algorithm can effectively maintain the consistency of the node sending rate,reduce the occurrence rate of network congestion and improve the quality of network services.?3?Aiming at the problem that the influence of network delay and noise on network congestion control in multi routing environment,the H_?active queue management algorithm applied to multi router is designed.Firstly,based on fixed network delay,the multi router network congestion control model is established.The consistency congestion control algorithm is proposed and is proved that the algorithm can satisfy the H_?performance.Secondly,considering the network delay variability and the delay bound determined by the TCP timeout retransmission mechanism,the active queue management algorithm is designed and its consistency and H_?performance is proved.The method is given to solve the parameters of the controller.Simulation results show that the proposed algorithm improves the stability and robustness.?4?Aiming at the problem that the communication delay affects the dynamical behavior of the TCP/AQM system in the wireless access network,the existence of Hopf behavior is proved,and the bifurcation controller is designed to control the bifurcation behavior effectively.Firstly,through analyzing the distribution of the root of wireless access network TCP/AQM model,it is proved that the system will appear Hopf bifurcation when the communication delay exceeds a critical value.The sufficient condition for the stability of system on the delaytis given.Secondly,based on the theory of nonlinear dynamics,the delay feedback controller is applied to control the Hopf bifurcation.The critical valuet₀?of Hopf bifurcation in relation to the controlled system is obtained.The stability range of the system is enlarged.At last,using the normal form theory and the center manifold theorem,the method to determine the direction of the Hopf bifurcation and the stability of periodic solutions refer to the controlled system is deduced.It reveals the influence of time delay on bifurcation properties of wireless access network TCP/AQM system.The simulation validates the correctness of the analysis and the effectiveness of the proposed control strategy.