Application of Least-Squares Channel Estimation to Large-scale MU-MIMO-OFDM in the Presence of Terminal Mobility

This paper investigates the effect of multiplexing-based least squares (LS) channel estimation on uplink large-scale MU-MIMO-OFDM in the presence of terminal mobility. In uplink large-scale MU-MIMO, the same number of pilot symbols as that of mobile stations (MSs) is generally required, which results in the consumption of a substantial amount of resources for channel estimation. This overhead can be reduced by using the LS channel estimation in user multiplexing because this method has only to use a limited number of subcarriers, with a multipath length at most. In addition, the achieved overhead reductions can be effectively utilized for improving the tracking capability of channel estimation under varying channels. Since our performance evaluation focuses on mobile communications systems such as 6G, uplink large-scale MU-MIMO is assumed, and Gaussian belief propagation (GaBP) is applied to detect multiple uplink signals. The effectiveness of this method relative to that of the traditional LS method with the same number of pilot symbols is demonstrated under time selective fading through computer simulations.