To relieve the stress on channel estimation and decoding complexity in cell-free massive multiple-input multiple-output (MIMO) systems, user grouping problem is investigated in this paper, where access points (APs) based on time-division duplex (TDD) are considered to serve users on different time resources and the same frequency resource. In addition, when quality of service (QoS) requirements are considered, widely-used max-min power control is no longer applicable. We derive the minimum power constraints under diverse QoS requirements considering user grouping. Based on the analysis, we formulate the joint power and user grouping problem under QoS constraints, aiming at minimizing the total transmit power. A generalized benders decomposition (GBD) based algorithm is proposed, where the primal problem and master problem are solved iteratively to approach the optimal solution. Simulation results demonstrate that by user grouping, the number of users served in cell-free MIMO systems can be as much as the number of APs without increasing the complexity of channel estimation and decoding. Furthermore, with the proposed user grouping strategy, the power consumption can be reduced by 2-3 dB compared with the reference user grouping strategy{, and by 7 dB compared with the total transmit power without grouping.
翻译:为了缓解无细胞的大型多投入多输出(MIMO)系统对频道估计和解码复杂性的压力,本文件调查了用户组别问题,认为基于时间分配的分解(TDD)的接入点为用户在不同时间资源和同一频率资源上提供服务;此外,在审议服务质量要求时,广泛使用的最大功率控制不再适用;在考虑用户群化的不同QOS要求下,我们根据不同的QOS要求得出最低功率限制。根据分析,我们在QOS限制下制定联合电力和用户组别问题,目的是最大限度地减少总传输力;提出了基于时间分配的通用弯曲器分解算法,其中原始问题和总问题通过迭接方式解决最佳解决办法;模拟结果显示,通过用户组别,在无细胞的MSIMO系统中服务的用户数量可以与AP数量一样多,而不会增加频道估计和解码的复杂程度。此外,在拟议的用户组别下,根据拟议的用户组别,将电力分位法(GB) 与总用户群比较,通过2至3的用户群比较,将电力与总消费战略相比,可以减少。