Research On Adaptive Queue Management Algorithm

With the rapid development of the Internet,multimedia applications have deeply changed people's living habits with the popularity of smart terminals.Whether it is work,study,social or entertainment,it is inseparable from the assistance of network communication.The development of network brings people great convenience but increases the risk of network congestion.Therefore,it is important to strengthen the research on network congestion management.In order to study the principle and method of network congestion management,the cause of congestion are analyzed firstly.And next,algorithms based on TCP and IP are studied.Finally,advantages and disadvantages of two congestion control mechanisms are analyzed and specific indicators for evaluating congestion management algorithm is given.In order to solve the problem of traditional queue management algorithm,a new queue management algorithm based on queue state model is proposed.The new algorithm forecasts the change trend of the queue size by network traffic and change trend of current queue size to adjust the discard function which greatly improves the response speed of the algorithm.The simulation results show that the new algorithm is more stable than the traditional queue management algorithm,and it has obvious improvement in performance indicators such as packet loss rate and throughput.In order to further improve the delay jitter performance of queue management algorithm based on queue state model.An adaptive active queue management algorithm based on queue state model is proposed.It optimizes the function of maximum dropping probability so that it can adapt to changes of network load.The simulation results show that this method obviously improve the delay jitter index as the network load increases.