基于802.11ac WLAN系统的比例公平调度算法研究

发布时间：2018-07-25 08:33

【摘要】：随着当今多媒体业务的爆炸式增长和人们对高速率传输网络的迫切需求,无线局域网得到迅速发展与普及。无线局域网协议802.11ac中,MU-MIMO(Multi-Users Multiple-Input Multiple-Output)传输可以大幅提高数据传输率速率,但是考虑到有限的无线资源与多媒体业务不断提高的服务质量要求之间的矛盾,如何在资源调度时平衡公平性和系统性能成为当今的研究热点问题之一。现有资源调度算法大多强调对无线资源的最佳利用,而忽视了用户服务质量和公平性,或者仅考虑分配的公平性和系统性能而没有考虑到用户实际业务的服务质量需求。针对上述问题,本文首次在MU-MIMO的流分配中同时考虑了用户服务质量、比例公平以及系统性能三个因素并提出了一种新的资源调度算法。本文所提算法不仅能为不同的业务提供相应的服务质量保证,而且可以有效提高资源调度的比例公平性而不会使系统的性能受到太大的影响。该算法在交叉熵算法的基础上引入松弛因子、剩余因子以及用户权值三个参数,在不同资源状态下,通过在迭代过程中实时调整以上三个参数的值来得到满足不同目标的资源调度结果。在无线资源不足的情况下,该算法的优化目标为使各用户所分资源达到其服务质量的相同比值,此时公平性是主要目标,提高系统性能是次要目标;而在资源充足的情况下,优化目标则变为在保证用户服务质量的前提下尽可能提高系统性能,此时提高系统性能是主要目标,用户间资源调度的公平性是次要目标。为了验证算法的有效性,本文将所提算法分别应用于确定性链路速率条件下单基站系统和多基站网络两种模型,仿真结果表明,该改进算法能实现用户间资源调度的公平性和系统吞吐量最大的良好折中。此外,本文针对时变链路速率网络,通过在已有的多基站协作的全局比例公平调度算法中引入松弛因子、剩余因子以及用户权值三个参数,并根据服务质量需要来实时调节资源的调度,从而实现用户时间平均吞吐量的比例公平。仿真表明,该改进算法能够得到满足比例公平的调度结果。
[Abstract]:With the explosive growth of multimedia services and the urgent need of high-speed transmission networks, WLAN has been rapidly developed and popularized. MU-MIMO (Multi-Users Multiple-Input Multiple-Output) transmission in WLAN protocol 802.11ac can greatly improve the rate of data transmission, but considering the contradiction between the limited wireless resources and the increasing quality of service requirements of multimedia services. How to balance fairness and system performance in resource scheduling has become one of the hot issues. Most of the existing resource scheduling algorithms emphasize the best use of wireless resources, but neglect the quality of service and fairness of users, or only consider the fairness of allocation and system performance without considering the QoS requirements of users' actual services. In order to solve the above problems, this paper firstly considers the three factors of user quality of service, proportional fairness and system performance in the flow allocation of MU-MIMO, and proposes a new resource scheduling algorithm. The proposed algorithm can not only guarantee the quality of service for different services, but also effectively improve the proportional fairness of resource scheduling without greatly affecting the performance of the system. On the basis of cross-entropy algorithm, the algorithm introduces three parameters: relaxation factor, residual factor and user weight. By adjusting the values of the above three parameters in real time during the iteration process, the resource scheduling results satisfying different objectives are obtained. In the case of lack of wireless resources, the optimization goal of the algorithm is to achieve the same ratio of quality of service (QoS) to the resources allocated by each user. At this time, fairness is the main goal, and the improvement of system performance is the secondary objective, while in the case of sufficient resources, the main objective is fairness. The optimization goal is to improve the system performance as much as possible under the premise of guaranteeing the user's quality of service. In this case, the main goal is to improve the system performance, and the fairness of resource scheduling between users is the secondary goal. In order to verify the validity of the algorithm, the proposed algorithm is applied to two models of deterministic link rate conditional sending order base station system and multi-base station network respectively. The simulation results show that, The improved algorithm can achieve a good compromise between the fairness of resource scheduling between users and the maximum system throughput. In addition, for time-varying link rate networks, by introducing relaxation factor, residual factor and user weight into the existing global proportional fair scheduling algorithm of multi-base station cooperation, this paper introduces three parameters: relaxation factor, residual factor and user weight. According to the need of quality of service, the scheduling of resources is adjusted in real time to realize the fair proportion of the average throughput of user time. Simulation results show that the improved algorithm can achieve proportional fair scheduling results.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN925.93

【相似文献】