【摘要】 该文考虑一种分布式大规模MIMO系统,假设基站端与用户之间的信道为莱斯信道,研究了该系统中基站选择的算法。首先给出了系统采用匹配滤波和迫零预编码时,用户下行可达速率的闭式表达式,并分析了系统的功率效率性能。然后基于此闭式表达式,以最大化系统的频谱效率为目标,提出了基于增量选择和基于用户优先级的基站选择算法。这两种算法只需要系统获取基站端与用户之间的信道统计特征信息,从而有效降低了系统开销。仿真结果表明,所提出的两种基站选择算法性能仍能逼近最优算法。特别地,当采用匹配滤波预编码且基站端天线数趋于无穷时,基于用户优先级的基站选择算法优于基于增量选择的算法。更多还原

【Abstract】 The Base Station(BS) selection for distributed massive MIMO system is studied. The channels between the BSs and the users are assumed to be Rician fading. First the closed-form expressions of downlink achievable rates for Matched-Filter(MF) and Zero-Forcing(ZF) precoding are derived respectively, and the power efficiency performance of the system is investigated. Based on the closed-form expressions, the approximate incremental BS selection and the approximate user priority BS selection algorithms are proposed, both of which only require the statistical Channel State Information(CSI), to maximize the system spectral efficiency. Numerical results show that the proposed algorithms perform close to the optimal one. Specifically, it is shown that the approximate user priority BS selection algorithm outperforms the approximate incremental one when the BSs utilize the MF precoding in the large scale antenna limit.更多还原