论文总字数:31358字
摘 要
04014449 韩磊鑫
指导教师 王家恒
随着移动互联网的迅速发展,人们对于不同业务数据需求持续增长,给当前的第四代移动通信系统带来了较大的压力。为了满足各种场景的通信需求,第五代移动通信系统(5G)提出的密集网络成为了目前研究的一个热点。
本文主要研究了在蜂窝系统中部署Small Cell的场景中的下行多输入单输出(Multiple-Input Single-Output, MISO)波束赋形问题,在保证宏蜂窝用户的最低通信速率也即服务质量(Quality of Service,QoS)的前提下,联合优化宏基站和小基站的波束赋形矢量,以最大化各基站的通信速率。以此建立了一个博弈优化问题,其约束条件耦合了所有基站的波束赋形向量,使得Small Cell网络资源分配优化问题变成了广义纳什均衡问题。本文首先通过迫零的方式消除小区内干扰,然后借助变分不等式理论给出了简化的变分不等式问题,并证明了变分均衡解的存在性和唯一性。最后本文设计了基于定价机制的分布式迭代算法,并证明了算法的收敛性。通过仿真表明该算法能收敛到变分均衡点并能够有效保证宏蜂窝用户的QoS要求。
其次,我还研究了Small Cell场景中下行链路的功率分配优化问题,目标函数也是在满足宏基站用户QoS约束的情况下最大化各基站的通信速率。类似地,利用变分不等式理论将该问题简化并采用基于最佳响应的分布式算法求解,最后通过仿真验证了算法的正确性。
关键词:密集网络,博弈论,广义纳什均衡,分布式算法
Abstract
04014449 Leixin Han
Advisor Jiaheng Wang
With the rapid development of the mobile Internet, people's requirement for different business data traffic continues to increase, which brings enormous pressure to the current fourth-generation mobile communication system. In order to meet the communication needs of various scenarios, the dense network proposed by the 5th generation mobile communication system (5G) has become a popular topic in the current research.
The Multiple-Input Single-Output (MISO) beamforming problem in the Small Cell scenario is studied in this paper. On the premise of guaranteeing quality of service (QoS) of macrocell users, joint optimization of beamforming vectors of macro base stations and small base stations is performed to maximize the communication rate of each base station. Based on the above, a game optimization problem is established. Because the constraint conditions couple the beamforming vectors of all base stations, the problem of resource allocation optimization for Small Cell networks becomes a generalized Nash equilibrium problem. Firstly, I eliminate the intra-cell interference by zero forcing, and then use the variational inequality theory to give a simplified variational inequality problem and prove the existence and uniqueness of the variational equilibrium solution. Finally, I designed a distributed iterative algorithm based on the pricing mechanism and proved the convergence of the algorithm. The algorithm can converge to the variational equilibrium point and can effectively guarantee the QoS requirements of macro cellular users through simulation.
In addition, I also studied the power allocation optimization problem of the downlink in the Small Cell scenario. Its objective function is to maximize the communication rate of each base station while satisfying the QoS constraints. Similarly, I use the variational inequality theory to simplify the problem and solve it with the distributed algorithm based on the best response. Finally, the correctness of the algorithm is verified by simulation.
KEY WORDS: Dense network,Game theory,Generalized Nash equilibrium,Distributed algorithm
数学符号说明
:矩阵
:向量
:单位矩阵
:矩阵的第i行第j列的元素
:矩阵的共轭转置
:矩阵的转置
:矩阵的秩
:矩阵的迹
:
:矩阵的最小特征值
:矩阵的谱半径
:向量的模
:
目 录
摘要 I
Abstract II
数学符号说明 III
第一章 绪论 1
1.1 研究背景 1
1.2 5G密集网络干扰协调研究现状 1
1.2.1 Small Cell研究现状简介 1
1.2.2 分布式网络优化介绍 2
1.3 论文章节安排 2
第二章 5G密集网络简介及相关理论基础 4
2.1 Small Cell简介 4
2.1.1 Small Cell分类 4
2.1.2 Small Cell和Macro Cell间的干扰分析 5
2.2 相关理论基础 6
2.2.1 凸优化简介 6
2.2.2 博弈论 7
2.2.3 变分不等式 8
2.3 本章小结 9
第三章 Small Cell下行链路中的分布式波束赋形优化 10
3.1 引言 10
3.2 系统模型与问题描述 10
3.3 问题转化 12
3.3.1 迫零方式消除小区内干扰 12
3.3.2 基于变分不等式的问题转化 14
3.4 分布式算法设计 17
3.4.1 均衡解的存在性和唯一性 17
3.4.2 求解的NE 18
3.4.3 价格因子更新 20
3.5 仿真结果及性能分析 21
3.6 本章小结 25
第四章 Small Cell下行链路中的分布式功率分配 26
4.1 引言 26
4.2 系统模型与问题描述 26
4.3 基于变分不等式的问题转化 27
4.3.1 GNEP与VI的等价转化 27
4.3.2 GNEP问题分解为NEP和GVI问题 27
4.4 分布式算法设计 28
4.5 仿真结果及性能分析 29
4.6 本章小结 31
第五章 全文总结与展望 32
5.1 论文内容总结 32
5.2 进一步的研究方向 32
致谢 33
参考文献 34
附录A 36
A.1 证明命题3.1和引理3.3 36
A.2 证明定理3.1和引理3.4 37
剩余内容已隐藏,请支付后下载全文,论文总字数:31358字
该课题毕业论文、开题报告、外文翻译、程序设计、图纸设计等资料可联系客服协助查找;