Research On The Network Control Algorithm Based On Active Queue Management

With the development of modern society, the Internet technology has entered a period of rapiddevelopment. The development of Internet makes the rapid and diversified growth of IP business.At the same time, along with the rising number of Internet users, the network congestion problem isbecoming more and more serious. So how to improve the network quality of service (QoS), andhow to manage the data in the network has become one of the hotspots in today’s society. In orderto satisfy the QoS performance indicators, congestion control methods play an important role inimproving the network congestion, and Active Queue Management (AQM) mechanism is also oneof the important methods that realize the congestion control.AQM becomes a focus in the field of network congestion control research and further study ofthe hot issues. There were a lot of improved algorithms based on basic AQM, this paper mainlydiscusses several commonly used AQM algorithms with deep analysis and comparison. In order toachieve fairness and ability to adapt the unity of the target, based on the enhanced gCHOKe fairness,the paper puts forward a new AQM algorithm based on pretreatment pgCHOKe. The algorithm ismainly based on the non-adaptive flows in the network bandwidth and packet numbers to detectnetwork congestion. When congestion found in the network will have a preprocessing filter towardsthe packets, and then send the packet to the gCHOKe algorithm. Ease the network congestion andreach the purpose of improving fairness and adaptability. At the same time using the algorithm ofquadratic function loss rate, improve the packet loss probability of the queue. NS2networksimulation results show that the algorithm not only can effectively detect congestion, adapt to thechange of flow, but also can effectively protect the response flow, improve the hit probability of theresponse flow, enhance the fairness and the stability of the queue length of the algorithm.