Antenna selection in massive MIMO systems: Full-array selection or subarray selection?

This paper presents a concept of subarray antenna selection in massive MIMO systems, i.e., the whole antenna array is segmented into subarrays and only one antenna is selected in each subarray. The subarray antenna selection can reduce the hardware complexity of the RF switching network, but at the cost of performance degradation due to the introduced selection constraint. We evaluate such performance degradation not only with the ideal independent and identically distributed (i.i.d.) but also with the realistic measured channels. To serve our evaluation, an antenna selection algorithm based on branch and bound method is presented, which aims to select an optimal channel submatrix with the largest minimum singular value. A qualitative analysis is also given to compare the performance between the full-array and subarray selection. Simulation results show that the massive MIMO subarray antenna selection has close performance to the full-array antenna selection, no matter with the i.i.d. channels or the realistic channels. The results in this paper imply that the with the concept of subarray antenna selection, antenna selection is still a practical and promising technology in the realistic massive MIMO systems.