Research About Network Congestion Control Based On Active Queue Management

With the high development of Internet, along with the rising number of Internet users and different applications, the requirement of users are higher and higher, the produced mass data and IP business are growing rapidly. Network congestion control play an important role in improving the network quality of service (QoS). And Active Queue Management (AQM) mechanism is one of the important technologies of congestion control which works on routers. AQM becomes a focus in the field of network congestion control research and further study of the hot issues in recent years. There were a lot of classic and mature AQM algorithms have been proposed.Several classic AQM algorithms are mainly analyzed with simulation in the thesis. In order to achieve better fairness of bandwidth and stronger ability to adapt the unity of the target, it proposed a new AQM algorithm sgCHOKe(Sampling based gCHOKe) based on sampling in the thesis. Through analyzing the low effectiveness of non-response flows hit on gCHOKe, which samples several packets from queue compare CHOKe with the arrived packet and employ a new mechanism of geometric distribution function. At the same time, The simulation results show that, sgCHOKe can maintain the fairness of different bandwidths better, achieve better effectiveness of non-response flows hit and guarantee fairness, also enhance the stability of the average queue length, at the same time, sgCHOKe has better performance on fairness index, packet loss probability and link utilization...