Geometry-Aided Joint Estimation of Short Range MIMO Channels with Hybrid Transceivers

In short-range millimeter wave (mmWave) multiple-input multiple-output (MIMO) communication scenarios the distance between transceivers is often comparable to the size of antenna arrays. The resulting line-of-sight (LoS) MIMO channel has typically a rank higher than one, and hence, conventional channel estimation methods based on the far-field approximation may not be applicable. In this paper, we consider the uplink channel estimation problem using a hybrid transceiver architecture and propose a geometry-assisted, maximum likelihood (ML)-based joint estimation of a short-range LoS channel with a reduced number of pilot transmissions. Under the near-field assumption, the proposed approach exploits the known RX geometry and describes the joint likelihood function based on the intersection of two antenna-specific reference angle of arrivals (AoAs). Relying on limited subarray-specific pilot measurements, all antenna-specific AoAs and subarray-specific correction factors can be computed using geometry dependencies among estimated reference AoAs, which in turn can be used for channel reconstruction. Simulation results show that the proposed method is able to reliably estimate the short range MIMO channel with a greatly reduced number of measurements compared to the previous works.