在C-RAN网络中基于稳定接入关系的基站激活算法研究

发布时间：2018-03-03 07:26

本文选题：无线云接入网络　切入点：基站激活　出处：《电子科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着移动端数据流量的爆炸式增长,传统的无线接入网络因为有效覆盖低,容量干扰受限,维护升级困难等原因而无法满足未来无线通信新要求。一种基于云架构的无线接入网络(Cloud Radio Access Network,C-RAN)应运而生。在C-RAN网络中,射频远头(Remote Radio Heads,RRH)可以密集部署并多点协作传输以增加信号质量和边缘覆盖率,但出于回传网络容量,干扰抑制,以及调度复杂度等方面的考虑,处于激活状态的RRH数量往往受到限制。在传统的RRH激活研究中,RRH的激活/休眠状态一般是基于信道的状态信息,考虑能耗,信道容量等因素进行选择。在这种选择方式中,系统因为匹配信道变化往往工作在一种高性能的状态,但当信道变化较快时,RRH状态也会快速变化,这会增加系统调度难度,加快设备损耗速度,甚至严重者会导致服务中断。针对这种情况本文重点研究了C-RAN网络中基于稳定接入关系的基站激活算法,旨在寻找一种能平衡系统灵活性和稳定性的调度策略。本文工作主要包括以下几个方面:(1)提出了一种基于基站用户稳定接入关系的资源调度和波束形成联合优化的数学模型。与传统网络模型不同,本模型充分考虑网络资源调度的灵活性与稳定性,不再局限于某一时刻系统的性能,而是着眼于网络长时间的资源调度分配策略以实现降低能耗提升容量的目的。(2)基于非因果信道设计了一种线下RRH激活策略,并在此基础上设计了两种基于因果信道的在线激活策略。系统长期性能模型不同于瞬时系统模型,它的求解往往受制于信道的因果性。本文先基于非因果信道探究了一种集中式的RRH激活算法,再在此基础上设计了两种针对因果信道的RRH激活策略:动态激活策略和训练式激活策略。仿真结果表明相较于传统的无切换控制激活算法,本文提出的三种RRH激活算法能够较好的提升系统的稳定性。(3)针对三种策略设计了基于交替方向乘子法(Alternating Direction Method of Multipliers,ADMM)的分布式求解方案。针对集中式求解方式中算法复杂度高的问题,提出了一种基于ADMM的分布式计算方法。ADMM算法能有效分离耦合的变量,将原问题分割为许多可以独立求解的子问题,不仅减小了计算复杂度,也十分契合C-RAN的网络架构。仿真结果表明,ADMM收敛性良好,并且能大幅度降低计算复杂度。
[Abstract]:With the explosive growth of mobile end data traffic, the traditional wireless access network is limited because of its low coverage and limited capacity interference. Because of the difficulty of maintaining and upgrading, it can not meet the new requirements of future wireless communication. A cloud-based wireless access network named Cloud Radio Access Network (C-RAN) has emerged. RF remote Radio head RH) can be deployed intensively and multipoint cooperative transmission to increase signal quality and edge coverage, but due to network capacity, interference suppression, scheduling complexity, etc. The number of RRH in active state is often limited. In the traditional research of RRH activation, the activation / dormancy state of RRH is usually based on the state information of the channel, considering the factors of energy consumption and channel capacity. The system usually works in a high performance state because of the matching channel change, but the RRH state also changes rapidly when the channel changes rapidly, which will increase the difficulty of system scheduling and speed up the equipment loss. In view of this situation, this paper focuses on the base station activation algorithm based on stable access relation in C-RAN network. In order to find a scheduling strategy that can balance the flexibility and stability of the system, this paper mainly includes the following aspects: 1) this paper proposes a joint optimization of resource scheduling and beamforming based on the stable access relationship of base station users. Different from the traditional network model, This model fully considers the flexibility and stability of network resource scheduling, and is no longer limited to the performance of the system at a certain time. Instead, an offline RRH activation strategy is designed based on the non-causal channel, which focuses on the resource scheduling and allocation strategy for a long time to achieve the purpose of reducing energy consumption and increasing capacity. On this basis, two online activation strategies based on causality channel are designed. The long-term performance model of the system is different from the instantaneous system model. Its solution is often constrained by the causality of the channel. This paper first explores a centralized RRH activation algorithm based on the non-causal channel. On this basis, two RRH activation strategies for causal channel are designed: dynamic activation strategy and training activation strategy. The simulation results show that compared with the traditional no-handover control activation algorithm, The three RRH activation algorithms proposed in this paper can improve the stability of the system. (3) for the three strategies, a distributed solution scheme based on alternating direction multiplier method and alternative direction multiplier method (Alternating Direction Method of Multipliers ADMMMMMMM) is designed. The complexity of the problem, This paper presents a distributed computing method based on ADMM. ADMM algorithm can effectively separate the coupled variables. The original problem can be divided into many sub-problems that can be solved independently, which not only reduces the computational complexity, but also reduces the computational complexity. The simulation results show that the convergence of ADMM is good and the computational complexity can be greatly reduced.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92;TN929.5

【相似文献】