面向5G的无线网络节能技术研究

发布时间：2018-01-11 05:31

本文关键词：面向5G的无线网络节能技术研究　出处：《东南大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着移动通信技术的快速发展和移动数据流量的指数式增长,超密集小蜂窝网络(SCN,Small Cell Network)与异构网络(Hetnets,Heterogeneous Networks)等新型网络的研究在业界得到了极大重视。以部署低功率接入点(LPN,Low Power Node)为特征的SCN拉近了用户设备(UE,User Equipment)与无线网络之间的距离,极大地提升了网络容量与用户体验,而宏微蜂窝重叠部署的Hetnets可以满足多样化的应用需求,这些新型网络架构已经成为构建5G移动通信系统的关键技术。然而,小蜂窝部署密度的增大在提升网络容量的同时,也带来了日益严重的网络能耗问题。因此,能够保证一定网络服务质量的网络节能技术得到了广泛关注。本论文针对面向5G的SCN与Hetnets网络节能技术展开研究。针对现有的无线网络节能技术进行充分的调研。首先围绕面向5G的无线网络的部署和架构特点,介绍了SCN和Hetnets。接着,分别重点研究了同构网络(Honets,Homogeneous Networks)与Hetnets的长时基站(BS,Base Station)节能技术,包括BS休眠与小区范围扩展。为实现Honets的节能目标,可以关闭低负载小区或扇区,剩余的开启小区或扇区通过覆盖范围扩展来补偿出现的覆盖漏洞。在此基础上,研究Hetnets节能技术,通常动态地让低负载LPN进入休眠模式,由MacroBS负责网络覆盖,让网络资源适应负载的空时波动性。此最后,简单介绍了无线网络的短时BS节能技术,以在更小的时间尺度上实现网络资源的最优配置。以降低超密集SCN的能耗为目标,研究网络负载动态变化条件下的Micro BS扇区休眠算法。针对三扇区Micro BS密集部署而构成的同构SCN场景,提出了周期性地动态配置所有扇区工作模式的两阶段扇区休眠算法。第一阶段算法依据扇区价值函数,按照合理次序对开启扇区进行测试,以尽可能多地关闭冗余扇区,有两种设计方案,其中启发式方案复杂度较高,对此,提出了渐进式方案,通过避免不必要的测试过程,克服启发式方案的不足。第二阶段扇区工作模式重调整算法让部分不适合休眠的扇区继续保持开启,并调整部分开启扇区的电下倾角,保障网络覆盖。接着,为了让移交UE重新选择合适的服务扇区与带宽资源,提出一个UE移交算法。仿真结果表明,所提出的休眠算法能够让开启的扇区数量有效地跟踪网络负载的波动,不同的第一阶段方案导致类似的节能效果,但渐进式方案可以显著减少休眠算法的执行时¨间。在上一章扇区休眠算法研究的基础上,继续研究Hetnets中微微基站(Pico BS)的休眠算法。从Hetnets系统模型出发,详细阐述了UE接入宏微BS的过程以及相应的无线资源分配方案。接着综合考虑Hetnets的特点,设计了Pico BS价值函数,以协助休眠算法按照合理顺序测试开启的PicoBS。随后提出了两种动态Pico BS休眠算法(即启发式与渐进式),通过周期性地配置Pico BS的工作模式,实现Hetnets的节能目标。启发式算法依赖价值函数逐个测试网络中所有的PicoBS,由于复杂度较高而不适合Pico BS密集部署的Hetnets。为了加速启发式算法的执行过程,提出了渐进式Pico BS休眠算法。对于被测试的Pico BS,为了尽可能地关闭Pico BS考虑首先将其UE移交到邻近宏基站(Macro BS),在移交不成功的情况下,再尝试将UE移交到邻近Pico BS。仿真结果表明,渐进式算法的执行时间平均只有启发式算法的三分之一,而且可以关闭更多的Pico BS但对阻塞概率只有轻微的不利影响,这意味着渐进式算法具有更好的节能潜力。
[Abstract]:With the rapid development of mobile communication technology and mobile data traffic, exponential growth, ultra dense small cellular networks (SCN, Small Cell Network (Hetnets) and Heterogeneous Networks, heterogeneous network) of the new network has gained much attention in the industry. For the deployment of low power access point (LPN, Low Power Node) features the SCN closer to the user equipment (UE, User, Equipment) and the distance between the wireless network, greatly enhance the network capacity and user experience, and macro micro cellular overlapping deployment of Hetnets can meet the diverse needs of the application, the new network architecture has become a key technology to build 5G mobile communication system. However, small cell the deployment density increases to improve network capacity at the same time, also brought the energy consumption of the network becomes a serious problem. Therefore, to ensure the energy saving technology of network service quality by network Aiming at the widespread concern. For 5G SCN and Hetnets network energy saving technology research. Full investigation in wireless networks, the existing energy-saving technology. The deployment and architecture of wireless network based on 5G firstly, then introduces SCN and Hetnets., respectively, focusing on the research of homogeneous networks (Honets, Homogeneous and Networks) Hetnets long base station (BS, Base Station) energy-saving technologies, including BS cell expansion and dormancy. In order to achieve energy saving goals of Honets, can turn off the low load cell or sector, the rest of the open cell or sector by covering the vulnerability coverage extended to compensation. On this basis, the research of Hetnets energy-saving technology, usually let the low load dynamically LPN into sleep mode, MacroBS is responsible for the network coverage, let the cyber source fluctuation load adaptive space-time. Finally, introduces the wireless network short BS energy-saving technology, to achieve optimal allocation of cyber source in smaller time scales. In order to reduce the energy consumption of super dense SCN as the target, the dynamic changes of Micro under the conditions of BS sector load research network dormancy algorithms. Based on the three sector of Micro BS consisting of densely deployed homogeneous SCN scene, the two stage cycle sector of all sectors of the working mode of the dynamic allocation algorithm. The first stage of dormancy algorithm based on sector value function, carries on the test to the open sector according to the reasonable order, to as much as possible to turn off redundant sectors, there are two kinds of design scheme, the heuristic scheme of high complexity, the proposed incremental scheme, by avoiding the testing process not necessary, overcome shortcomings of heuristic scheme. The sector working mode second phase adjustment algorithm is not suitable for some heavy dormant sector continues to remain open, and adjust the fan from the start The electronic angle, to ensure network coverage. Then, in order to allow the transfer of UE to select the appropriate service sector and bandwidth resources, and proposed a UE handover algorithm. The simulation results show that the proposed algorithm can make the dormant open sectors to effectively track the network load fluctuations, the first phase of the program lead to different energy saving effect similar, but the incremental scheme can significantly reduce the execution time to sleep algorithm. Based on the basic dormancy in the last chapter sector, continue to study Hetnets (Pico BS) of the pico base station sleep algorithm. Starting from the Hetnets system model, describes the process of UE access micro macro BS and wireless resources the corresponding allocation scheme. Then considering the characteristics of Hetnets, design the Pico BS value function, to help open the sleep algorithm according to a reasonable sequence test PicoBS. then put forward two kinds of dynamic Pico BS Hugh Sleep algorithm (i.e., heuristic and progressive), through periodic configuration Pico BS working mode, energy saving can be achieved Hetnets. All the heuristic algorithms rely on the PicoBS test network value function one by one, because of high complexity and is not suitable for Pico BS Hetnets. in order to carry out intensive deployment to accelerate the process of heuristic algorithm, we propose the progressive Pico BS sleep algorithm. For the tested Pico BS, in order to close the Pico BS UE will be the first consideration is handed over to the nearby Acer station (Macro BS), in the transfer of the unsuccessful cases, try to move UE to the adjacent Pico BS. simulation results show that 1/3 of the incremental algorithm execution time the average only heuristic algorithm, but also can close more Pico BS but on the blocking probability of only minor adverse effects, which means that the energy saving potential of the incremental algorithm has better.

【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN929.5

【相似文献】