QoS保障的基站节能机制研究

发布时间：2018-06-16 13:21

本文选题：能量效率 + 节能　；参考：《中国科学技术大学》2014年博士论文

【摘要】：随着移动互联网的发展,无线业务出现爆炸性增长。这种增长不仅体现在数据量的增加,也体现在业务种类和QOS需求的多样性变化。为了满足业务需求、提升网络容量,需要从传输技术和网络架构两个方面进行下一代无线网络设计。先进的物理层技术,如OFDM、MIMO,在增大频谱效率的同时,带来了更大的能量开销。未来网络站点密集化、异构化不仅增加站点数目、增加耗能,还提升网络干扰强度,降低了整网的能量效率。为了降低能量开销,减少碳排放,需要在保证业务需求的前提下对基站进行节能设计。本文先从业务QOS的表征参数出发,将时延作为最重要的QOS影响因素,并按照业务的时延要求,将业务分为严格时延约束和时延容忍两大类。针对这两类业务,分别从链路级角度设计时延保障的节能调度器,并考虑实际多小区系统,设计分层调度器,在保证时延要求的前提下最小化基站耗能,并协调相邻小区间休眠子帧,以降低网络间干扰。所有的这些研究都是基于实际的功耗模型进行的。由于基站功耗模型非连续,调度问题在多数场景下没有最优解。本文研究了严格时延约束业务的高能效调度策略。首先针对AWGN信道,研究如何在给定时间内传输不同时延要求的数据包使得系统所需的能量最小的问题,利用拟凸优化提出了低复杂度最优传输速率和休眠控制策略,在此基础上,提出因果系统的高能效传输调度器设计；进一步需要考虑衰落信道下存在信道估计耗能时的数据包级能效最优传输控制和调度策略,通过联合考虑信道的统计特性、导频耗能和数据包长度,提出包含剩余时间项、信道相关项和能量相关项的次优调度算法进行能效资源调度。本文针对时延容忍业务进行能效传输设计,提出两种休眠策略并利用排队论中带休假的M/G/1模型,计算出不同休眠控制策略下的能耗和业务平均时延的闭式表达,利用凸优化理论和拉格朗日乘子法对问题进行迭代求解,并提出给定拉格朗日乘子下的迭代凸搜索的最优功率分配和休眠控制算法；在多用户场景下,分析ARQ对时延影响,依据业务的统计特性建立优化问题,利用组合优化理论进行能效资源分配,在保证业务平均时延要求下最大化系统能效。本文最后研究实际多小区中满足用户QoS需求的子帧级协调休眠调度策略,提出两层调度器模型,利用线性控制理论设计上层调度器的等效滤波器,为每帧分配数据以保证QoS,底层进行高能效资源分配；先针对单用户,将活跃子帧连续化,利用拟凸优化证明问题存在最优解,通过迭代搜索求得最优功率和休眠子帧数目；然后考虑多用户系统,利用时间共享条件和拟凸优化可以获得休眠子帧数目,利用边缘自适应算法分配子信道、功率；最后对于多小区,依据单小区的资源分配结果确定各小区在每帧内所需传输数据量,依据上一时隙信道状态和干扰强度分布结果,动态协调休眠子帧,使得强干扰邻小区休眠子帧正交,进而提升网络能量效率。
[Abstract]:With the development of mobile Internet, wireless services have exploded. This growth is not only reflected in the increase in the amount of data, but also in the variety of business types and QOS requirements. In order to meet business needs and improve the network capacity, the next generation of wireless networks should be designed from two aspects of transmission technology and network architecture. The physical layer technology, such as OFDM, MIMO, increases the energy cost while increasing the spectrum efficiency. In the future, the network site intensive, isomerization not only increases the number of sites, increases energy consumption, but also improves the network interference intensity and reduces the energy efficiency of the whole network. In order to reduce the energy cost and reduce the carbon emissions, it is necessary to ensure the business needs. Under the premise, the energy saving design of the base station is carried out.
In this paper, starting from the characterization parameters of the service QOS, the time delay is considered as the most important QOS influence factor, and the service is divided into two categories, strict delay constraint and delay tolerance according to the time delay requirements of the business. The layer scheduler minimizes the energy consumption of the base station and coordinates intercell dormant subframes between adjacent cells to reduce internetwork interference. All these studies are based on the actual power model. The scheduling problem has no optimal solution in multiple scenes because of the discontinuous power model of the base station.
In this paper, the high-energy efficient scheduling strategy of strict time delay constrained service is studied. Firstly, the problem of how to transmit the required data in a given time in a given time is studied for the AWGN channel to minimize the energy required for the system. By using quasi convex optimization, a low complexity optimal transmission speed rate and a dormancy control strategy are proposed. The high energy efficiency transmission scheduler design of the causality system; further need to consider the packet level energy efficiency optimal transmission control and scheduling strategy in the channel estimation energy dissipation under the fading channel. By considering the statistical characteristics of the channel, the pilot energy consumption and the packet length, the residual time term, the channel correlation and the energy correlation term are proposed. The suboptimal scheduling algorithm is used for energy efficient resource scheduling.
In this paper, two kinds of sleep strategies are designed for time delay tolerance service. Two kinds of dormancy strategies are proposed and the closed expression of energy consumption and service average delay under different sleep control strategies is calculated by using the M/G/1 model with vacation in queuing theory. The convex optimization theory and Lagrange multiplier method are used to solve the problem iteratively, and the given pull is given. The optimal power allocation and dormancy control algorithm for iterative convex search under the Grange multiplier is used. Under multi user scenarios, the effect of ARQ on time delay is analyzed, and the optimization problem is established according to the statistical characteristics of the service. The energy efficiency resource allocation is carried out by combined optimization theory, and the energy efficiency of the system is maximized under the guarantee of the service average time delay.
At the end of this paper, a sub frame coordinated dormancy scheduling strategy which satisfies the user's QoS needs is studied. A two layer scheduler model is proposed and the equivalent filter of the upper level scheduler is designed by using linear control theory. The allocation of high energy efficiency resources at the bottom of each frame is guaranteed by QoS, and the active subframe is continuous for the single user. By using pseudo convex optimization, the optimal solution is proved, the optimal power and the number of dormant subframes are obtained by iterative search, and then the number of dormant subframes can be obtained by using the time sharing condition and pseudo convex optimization, and the sub channel and power are allocated by the edge adaptive algorithm. Finally, the multiple cell, based on the single cell, is based on the single cell. The result of resource allocation determines the amount of data that each cell needs to transmit in each frame. According to the results of the state of the slots channel and the distribution of the interference intensity, the dormant subframe is dynamically coordinated so that the strong interference neighbour subframe is orthogonal, and then the energy efficiency of the network is improved.
【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN929.5

【共引文献】