异构融合网络环境下负载均衡算法研究

发布时间：2018-03-04 15:39

本文选题：异构融合网络　切入点：负载均衡　出处：《东北大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着无线通信技术的发展和新的无线接入技术的出现,异构无线网络融合成为下一代无线通信系统的重要特征之一。负载均衡作为异构融合网络无线资源管理的关键技术,能够提高系统的无线资源利用率,有效避免系统性能恶化,提升系统服务质量,因此对负载均衡技术的研究具有十分重要的理论价值和实际意义。本文首先介绍了异构融合网络的基本理论,阐述了负载均衡的三种关键技术,深入研究了异构融合网络环境下的负载均衡算法,并对现有算法的优缺点进行分析和比较,在此基础上提出了有效的适于实际应用的负载均衡算法。针对现有负载均衡算法在准入控制方面存在滞后性和被动性的问题,提出了基于预测的动态负载均衡算法。利用互信息理论对异构融合系统中各个网络的负载状况进行预测,再依据所预测的结果最小化各个网络之间负载率方差,建立业务准入的优化控制模型,并通过遗传算法求得最优方案。所提算法能够降低RT业务平均阻塞率,减少NRT业务平均传输时间,提高准入控制的实时性和主动性,提高系统服务质量。仿真实验表明,与传统算法相比,基于预测的动态负载均衡算法能够高效利用无线网络资源,改善系统负载均衡水平。针对现有负载均衡算法在负载转移方面存在着频繁切换问题,提出了基于综合效用的负载均衡算法。通过构造适用于不同业务的效用函数,并利用三角模算子对不同效用函数进行融合得到候选网络的综合效用函数,提高了目标网络的综合评估性能,利用辅助动态转移因子制定合理的负载业务转移量,将重载网络中的负载业务通过垂直切换转移到轻载网络中,有效地减少了业务切换的次数,提高了资源利用率,改善了通信系统的传输性能。仿真实验验证了算法的有效性和优越性。
[Abstract]:With the development of wireless communication technology and the emergence of new wireless access technology, heterogeneous wireless network convergence has become one of the important features of the next generation wireless communication system. Load balancing is the key technology of heterogeneous convergence network wireless resource management. It can improve the utilization rate of wireless resources, effectively avoid the deterioration of system performance, and improve the quality of service of the system. Therefore, the research of load balancing technology has very important theoretical value and practical significance. Firstly, this paper introduces the basic theory of heterogeneous fusion network, and expounds three key technologies of load balancing. In this paper, the load balancing algorithm in heterogeneous fusion network environment is deeply studied, and the advantages and disadvantages of the existing algorithms are analyzed and compared. On the basis of this, an effective load balancing algorithm suitable for practical application is put forward, aiming at the problem of lag and passivity in the access control of the existing load balancing algorithm. A dynamic load balancing algorithm based on prediction is proposed. The load condition of each network in heterogeneous fusion system is predicted by mutual information theory, and the variance of load rate is minimized according to the predicted results. The optimal control model of service access is established, and the optimal scheme is obtained by genetic algorithm. The proposed algorithm can reduce the average blocking rate of RT services, reduce the average transmission time of NRT services, and improve the real-time and initiative of admission control. Simulation results show that compared with traditional algorithms, dynamic load balancing algorithm based on prediction can make efficient use of wireless network resources. In order to solve the problem of frequent switching between existing load balancing algorithms in load transfer, a load balancing algorithm based on comprehensive utility is proposed. The comprehensive utility function of candidate networks is obtained by combining different utility functions with triangular modular operators, which improves the comprehensive evaluation performance of the target networks, and makes use of auxiliary dynamic transfer factors to formulate a reasonable load traffic transfer amount. The load service in the heavy load network is transferred to the light load network by vertical switching, which effectively reduces the number of service switching and improves the resource utilization. The transmission performance of the communication system is improved and the effectiveness and superiority of the algorithm are verified by simulation experiments.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN92

【相似文献】