| 徐顺清* ** *** ****,石晶林* ** ***,周一青* ** ***,张宗帅* ** ***.基于波束训练的大规模 MIMO 多用户度调度算法[J].高技术通讯(中文),2022,32(3):248~260 |
| 基于波束训练的大规模 MIMO 多用户度调度算法 |
| Large-scale MIMO multi-user scheduling algorithm based on beam training |
| |
| DOI：10.3772/j.issn.1002-0470.2022.03.004 |
| 中文关键词: 大规模多输入多输出（MIMO）； 混合连接结构； 波束搜索； 用户选择； 频谱效率 |
| 英文关键词: massive multiple-input multiple-output ( MIMO), hybridly-connected structure, beam search, user select, spectrum efficiency |
| 基金项目: |
| 作者 | 单位 | | 徐顺清* ** *** **** | | | 石晶林* ** *** | | | 周一青* ** *** | | | 张宗帅* ** *** | |
|
| 摘要点击次数: 1293 |
| 全文下载次数: 1905 |
| 中文摘要: |
| 在毫米波多用户大规模天线阵列多输入多输出（MIMO）系统中，由于使用混合波束形成结构，使得用户调度问题面临新的挑战，调度问题转变为在同一时频资源选择调度用户子集和用户与基站之间通信波束对的问题，需要从用户和波束两个维度来进行考虑。大规模MIMO由于在基站端和用户端使用大量的天线阵列，如果使用穷举搜索算法运算量极大，在实际应用中是不现实的。本文提出了一种基于波束训练的多用户调度方案，该方案分为2步。第1步，由于波束间的干扰大多数来自于相邻波束，因此对相邻波束进行分组，同时根据用户反馈的信干燥比(SINR)最大值的基站波束索引号把相关的用户映射到对应波束组。第2步，为了减小干扰，每个波束组的最小波束索引根据一定的原则来选取调度用户，然后通过基站和用户之间的波束训练选取最接近于SINR门限值的用户波束来进行通信。仿真结果表明，该方案具有逼近最优穷举方案的性能并且具有较低的计算复杂度。 |
| 英文摘要: |
| In the millimeter wave multi-user massive multiple-input multiple-output ( MIMO) system, the structure is formed by the use of hybrid subdivision, so that the user scheduling problem faces new challenges. The scheduling problem is transformed into the problem of selecting scheduling user subsets and communication beam pairs between users with stations at the same time-frequency resource. The problem of two-way communication between substrates needs to be considered from the two dimensions of user and duality. Massive MIMO uses a large number of antenna arrays at the base station and the user terminal, and if exhaustive search algorithms are used, the computational complexity is extremely large, which is unrealistic in practical applications. A multi-user scheduling scheme based on beam training is proposed in this paper, which is divided into two steps. The first step is to group the adjacent beams and at the same time map related users to corresponding beam groups according to the beam index number of the maximum signal to interference noise ratio ( SINR) feedback from the user since most of the interference be-tween beams comes from adjacent beams. The second step is to reduce interference from the minimum beam index of each beam group to select scheduling users according to certain principles, and then select the user beam closest to the SINR threshold for communication through beam training between the base station and the user. The simula-tion results show that this scheme has the performance of approaching the optimal exhaustive scheme and has low computational complexity. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
