Research On Proportional Fair Scheduling Algorithm For 802.11ac WLAN System

With the explosive growth of multimedia service and the urgent demand for the high speed transmission network of people, the WLAN(Wireless Local Area Networks) has obtained the fast development. In the protocol of 802.11 ac, MU-MIMO(Multi-Users Multiple-Input Multiple-Output) can significantly enhances transmission speed. While, given the contradictions between the limited wireless resources and the increasing requirement of multimedia service, how to achieve a balance between user fairness and system performance in resource scheduling has become one of the research hotspots.The existing algorithms of wireless resource scheduling only emphasize the best use of the wireless resources, while either neglecting both of the Qo S(Quality of Service) and userfairness, or merely focusing on the fairness of resource scheduling and system performance but ignoring the Qo S of users. In terms of the defects above, this paper takes Qo S, proportional fair and system performance into consideration simultaneously when allocating spatial streams in MU-MIMO transmission for the first time, and puts forward a new resource scheduling algorithm.The proposed algorithm can not only satisfy corresponding Qo S for different business, but also effectively increase the proportional fair of resource scheduling with light decrease of system performance. Slack variable, surplus variable and user weight are introduced into resource scheduling algorithm based on cross-entropy method. By real-time adjustment of the above three parameters in iterative process, the proposed algorithm can satisfy different resource allocation target under corresponding situations. First, in the case of insufficient resources, the primary optimization goal is to reach the same ratio of the actual assigned resources and Qo S, the secondary is to improve the system throughput. Second, in the case of sufficient resources, the primary optimization goal is to improve system performance on the premise of guarantee the Qo S and the proportional fair allocation of radio resource is secondary. In order to verify the validity of the proposed algorithm, the arithmetic is applied to data center and multi-base station network with deterministic link rate. Simulations illustrate that the algorithm can make a compromise between proportional fair and system performance.Except for the proposed proportional fair scheduling algorithm under the condition of deterministic link rate, this paper ameliorates the GPF-BC(Global Proportional Fair scheduling-Base station Cooperation) method by introducing Slack variable, surplus variable and user weight into GPF-BC. The improved GPF-BC algorithm can rectify iterative process based on Qo S. Simulations illustrate that the scheduling result of improved algorithm can meet the requirements of proportional fair.