Low-Overhead Iterative Channel Parameter Estimation for Multi-User OAM Wireless Backhaul

Abstract

This paper considers the issue of acquiring channel state information (CSI) for multi-user orbital angular momentum (MU-OAM) wireless backhaul between the macro base station (MBS) and small base stations (SBSs) within broadcasting networks. Unlike prior works, we assume that each SBS transmits a pilot signal of length one on each multiplexed OAM mode and subcarrier, resulting in the coherent observations collected at the MBS. Then, we construct the data sets using the coherent observations, the components of which independently contain arbitrarily assumed positional information. The amplitude-phase multiple signal classification (AP-MUSIC) algorithm, a novel variant of the MUSIC, then conducts a two-dimensional (2-D) search on the amplitude and phase of the data component in both the OAM mode and frequency domains for estimating positions at each iteration. These estimates, together with the observations, are used to iteratively update the data sets, ultimately refining the distances and AoAs of all SBSs. The theoretical analysis and simulation results indicate that this solution not only yields the precise CSI for the MU-OAM system, but also markedly reduces the training overhead, compared to existing alternatives.