基于能效优化的异构融合网络智能调度技术研究

发布时间：2018-03-19 15:02

本文选题：异构融合网络　切入点：能量效率　出处：《西安电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着移动互联网的普及,为了解决移动信息业务数据量快速增长与无线网络传输容量受限之间的矛盾,网络融合技术作为一种有效途径,受到越来越多的关注与研究。此外,从绿色通信的角度,能量效率已经成为诸多无线网络重要的性能指标之一。本文针对多媒体内容的智能调度,应用随机几何理论和泊松点过程对无线通信网络中的基站和用户进行建模和分析,重点研究广播蜂窝异构融合网络中的能量效率问题。为了获得更高的网络吞吐量和能量效率,利用现有数字广播网络改造后所释放的UHF频谱资源,实现微蜂窝网络和数字广播网络的异构融合,从而为无线移动宽带需求的快速增长提供最为有效的技术手段。在融合网络中,定义需要传输的多媒体内容总数为M和内容调度中心调度给广播网络的多媒体内容的比例为per(或者数目n),对用户请求的多媒体内容进行智能调度,基本的调度策略为:按照多媒体内容的请求概率由大到小进行排序,概率较高的??M*per??或者n种多媒体内容由数字广播网络来传输,剩余的由蜂窝网络来传输。对融合网络的研究分为以下两方面:1.从用户请求多媒体内容的角度出发,根据用户请求多媒体内容的概率在Zipf分布中的排序,将请求的内容分为公共内容和个性化内容两类。基于随机几何空间网络模型,考虑广播网络的覆盖范围、蜂窝网络基站的激活概率和用户的中断概率,推导广播蜂窝异构融合网络的有效容量,并计算融合网络的能量效率。本文将per作为参考变量,分析该变量对融合网络能效的影响。仿真结果表明,当调度比例per给定时,在融合网络中微蜂窝基站的部署密度小于用户密度的情况下,融合网络的能效高于同等参数配置下单蜂窝网络的能效。2.由于第三章中融合网络能量效率表达式过于复杂,难以从理论上给出per的最优值,为此,本文对该能量效率表达式进行分析与化简。以简化场景下融合网络能量消耗为优化目标,以广播网络传输的多媒体内容数和带宽分配因子为约束条件,建立优化问题,并通过引入Lambert W函数对该优化问题进行求解。仿真结果表明,随着多媒体内容总数的增加,广播网络传输的多媒体内容数和带宽分配因子的仿真曲线逼近理论曲线,从而验证了优化问题求解结果的正确性。
[Abstract]:With the popularity of mobile Internet, in order to solve the contradiction between the rapid growth of mobile information services and the limitation of wireless network transmission capacity, network convergence technology, as an effective way, has been paid more and more attention and research. From the point of view of green communication, energy efficiency has become one of the important performance indicators of wireless networks. Using stochastic geometry theory and Poisson point process to model and analyze the base stations and users in wireless communication networks, the energy efficiency problem in broadcast cellular heterogeneous fusion networks is studied in order to achieve higher throughput and energy efficiency. Using the UHF spectrum resources released by the existing digital broadcasting network, the heterogeneous integration of microcellular network and digital broadcast network is realized. In order to provide the most effective technical means for the rapid growth of the demand for wireless mobile broadband. Defines that the total number of multimedia content to be transmitted is M and the proportion of multimedia content scheduled by the content scheduling center to the broadcast network is per( or a number of nngs) to intelligently schedule the multimedia content requested by the user. The basic scheduling strategy is: sort according to the request probability of multimedia content from large to small, the probability is higher? ? Mimper? ? Or n kinds of multimedia content are transmitted by the digital broadcast network, and the rest is transmitted by the cellular network. The research on the fusion network is divided into the following two aspects: 1.From the point of view of the user requesting the multimedia content, According to the ranking of the probability of users requesting multimedia content in the Zipf distribution, the requested content is divided into two categories: public content and personalized content. Based on the stochastic geometric space network model, the coverage of broadcast network is considered. The activation probability of the base station and the interruption probability of the user in the cellular network are derived, the effective capacity of the broadcast cellular heterogeneous fusion network is derived, and the energy efficiency of the fusion network is calculated. In this paper, per is used as the reference variable. The effect of this variable on the energy efficiency of fusion network is analyzed. The simulation results show that when the scheduling ratio per is timed, the deployment density of the microcellular base station in the fusion network is less than the user density. The energy efficiency of the fusion network is higher than that of the cellular network with the same parameters. 2. Because the expression of the energy efficiency of the fusion network is too complicated in Chapter 3, it is difficult to give the optimal value of per theoretically. In this paper, the energy efficiency expression is analyzed and simplified. In order to optimize the energy consumption of the fusion network under the simplified scenario, the optimization problem is established with the number of multimedia content transmitted by the broadcast network and the bandwidth allocation factor as the constraint conditions. The Lambert W function is introduced to solve the optimization problem. The simulation results show that with the increase of the total number of multimedia content, the simulation curve of multimedia content number and bandwidth allocation factor of broadcast network approximates to the theoretical curve. Thus, the correctness of the solution results of the optimization problem is verified.
【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】：2014
【分类号】：TN929.5;TP393.01

【相似文献】