Based On The Active Queue Management Algorithm Of Network Congestion Control Research

With the development of computer network, the comp uter network has been promoted theprocess of the world information. More and more new applications are launched to satisfy theneeds of people, which more and more people grow demand for network and network quality ofservice requirements have become more demanding. The rapid growth of Internet data flowmake the network appeared congestion phenomenon, resulting in network degradation, packetloss rate of the intermediate nodes and the network delay increases, the service of the user lackof timely response, to address the network congestion has become a very important issue. On theone hand, Network researchers need study how to use and integrate the existing networkresources to enable the network to achieve the highest performance; On the other hand, theyshould study new network protocols and algorithms to make a prospective study for thedevelopment of network.The primary cause of the network to produce congestion lies in the load provided by thenetwork is greater than the capacity and processing power of network resources. TCP congestioncontrol mechanism based on source nodes is effective in the prevention of congestion collapse,when congestion occurs the TCP source should reduce the transmission speed. But only usingTCP congestion control mechanisms cannot assure the robustness of Internet any longer, instead,the router should actively participate in the network congestion control for the most effectivedetection of congestion can occur in the router itself. Active Queue Management mechanism isan important part of congestion control, it can effectively monitor the length of the queue. Thispaper focuses on the stability and fairness of network, and proposes the solution aimed at REDalgorithm and CHOKe algorithm.(1) This paper analyses the relational expression between average queue size and packetloss rate. Based on the phenomena of Random Early Detection (RED) drops packet ratio is highwhile the network congestion is not serious and vice versa. This paper, in order to solve theabove problems, puts forward a non-linear high order congestion to control the ineffectiveness ofthe network. The arithmetic has established a high orde r function model which has goodcongestion control ability. By using the non-linear control to mark or discard fragments withlower and higher probability near low and high threshold respectively, it can control averagequeue length effectively. Meanwhile, NS2stimulation already verified that the arithmetic iseffective in improving the network performance.(2) This paper focuses on several kinds of fair AQM algorithms such as CHOKe and someother improved algorihms. In order to solve the problem of CHOKe algorithm cannot punishingthe non-responsive flows enough and the low level of accuracy, a new algorithm, LRU-CHOKe to penalty for non-responsive flows, is proposed in the paper. The algorithm is not only replacethe CHOKe hit with LRU hit to improve the effectiveness of CHOKe hit, but also use Queue hitto adaptively determine the number of packet loss. A new way of dropping packets to punishnon-responsive flows is adopted. The simulation results show that LRU-CHOKe is performedbetter than CHOKe in punishing non-responsive flows. As a result, the bandwidth allocation isrealized more fairly.