Hybrid Precoding Design for Adaptive Subconnected Structures in Millimeter-Wave MIMO Systems

Hybrid precoding design for an adaptive subconnected structure offers a compromise between hardware complexity and spectral efficiency in massive multiple-input multiple-output (MIMO) systems at millimeter-wave (mmWave) frequencies. This paper focuses on the impacts of the partial antenna connections on the achievable sum-rate (ASR) performance for the uniform linear array and uniform square planar array configurations in mmWave networks. The subarray architectures consist of the overlapped, interlaced, and dynamically selective subset schemes in mmWave MIMO systems. Notably, on the basis of the ASR maximization, the Gram-Schmidt (GS)-based antenna selection (AS) algorithm is used to acquire an appropriate antenna subset and magnificently simplifies the radio frequency chain-antenna mapping operation in the sequel. Finally, simulation results demonstrate that an adaptive subarray architecture in conjunction with the GS-based AS search strategy is able to accomplish the ASR performance approaching the fully connected architecture with an alleviated demand on both computation and hardware complexities.