Sensing-Assisted Beam Tracking for mmWave V2I Communications With Analog, Hybrid, and Digital Antenna Architectures

Abstract

In this paper, we study the sensing-assisted beam tracking (BT) problem in millimeter wave (mmWave) multiple-input multiple-output (MIMO) communications with different constraints on the antenna feeding networks. In particular, we consider a monostatic setup where a multi-antenna base station (BS) transmits information bearing orthogonal frequency division multiplexing (OFDM) signals via narrow beams, and a sensing receiver, co-located with the BS, listens to the echoes and estimates the adaptive steering needed in the BS transmit beams. Motivated by cost and space limitations of wide bandwidth mixed signal components, we propose and evaluate estimation and tracking algorithms for analog, digital, and hybrid digital analog (HDA) antenna architectures. The analog tracker is based on a novel learned variation of the hidden Markov model (HMM) filter, whereas HDA and digital trackers are based on much simpler methods, given their much richer corresponding measurements. Our results show that analog architectures attain performance close to that of fully digital setups while keeping system complexity much lower. All trackers are also shown to be capable of operating in multi user scenarios. The effect of transmit power and interval between measurements for the different architectures is thoroughly discussed in our numerical results.