Distance-Aware Subarray Selection for Terahertz Ultra-Massive MIMO Systems

As a means to support extremely high data rates in 6G wireless networks, terahertz (THz) ultra-massive multiple-input multiple-output (UM-MIMO) systems have attracted great interest in recent years. Unfortunately, due to the strong directivity and severe attenuation of the THz signal, the number of propagation paths is at most a few in the THz band. In most cases, therefore, the THz channel matrix is a low-rank matrix, which dramatically limits the channel capacity of THz systems. To increase the channel capacity of THz systems, the array-of-subarray (AoSA) technique that exploits a group of widely-spaced antenna subarrays has been proposed. A major issue of the AoSA scheme is that the base station (BS) has to employ a large number of subarrays along with the radio frequency (RF) chains connected to the subarrays so that the power consumption is considerable. In this paper, we propose an efficient THz UM-MIMO subarray architecture maximizing the channel capacity while reducing the power consumption. Key idea of the proposed scheme referred to as distance-aware subarray selection (DSS), is to choose a small number of subarrays maximizing the channel capacity, and then activate only the RF chains connected to the chosen subarrays. From the simulation results, we demonstrate that the proposed DSS scheme achieves a significant channel capacity gain over the conventional schemes.