Performance Analysis of Cell-Free Massive MIMO System With Limited Fronthaul Capacity and Hardware Impairments

摘要

We study uplink of cell-free massive multiple-input multiple-output system with limited capacity fronthaul link connecting each access point (AP) to a central processing unit (CPU), under hardware impairments at user equipments and APs. Achievable rates are derived for three transmission strategies at APs, named as compress-forward-estimate (CFE), estimate-compress-forward (ECF), and estimate-multiply-compress-forward (EMCF), to efficiently share the fronthaul capacity for sending channel state information (CSI) and/or data signals of the users to the CPU. For forwarding the compressed version of the signals, low-complexity fronthaul rate allocations are proposed for ECF and EMCF strategies, which considerably improve the performance especially for limited capacity fronthauls. Besides, for CFE and ECF strategies, two power allocations are presented to maximize the sum rates; One approximate solution based on geometric programming and an alternative general iterative algorithm. It is shown that performances of the strategies are not the same, because they require different fronthaul rate allocations for CSI and/or data signals transmission to CPU, and their AP signal processing capabilities are different. Finally, to highlight performance characteristics of the proposed system under different transmission strategies and circumstances such as pilot reuse and multiple-antenna APs scenarios, numerical results are presented and the three strategies are compared.