Multi-User Wideband Sparse Channel Estimation for Aerial BS with Hybrid Full-Dimensional MIMO

Due to the air-ground millimeter-wave channels exhibit inherent sparsity in both angle and delay domains, we propose a multi-user wideband sparse channel estimation scheme for aerial base station (BS) with hybrid full-dimensional (FD)-MIMO systems, where the aerial BS employs FD-MIMO array to serve multiple user equipments with a single antenna through line-of-sight (LoS) paths. Specifically, we first formulate multi-user uplink channel estimation as a parameter estimation problem. Then, by designing the hybrid combiners at the BS to visualize hybrid analog/digital array as a digital array, we estimate the horizontal/vertical AoAs and delays using the 3-dimensional (3-D) unitary ESPRIT algorithm. Moreover, the associated path gains are obtained using the least squares (LS) estimator. Finally, we reconstruct the MIMO channel based on the dominant channel parameters estimated above. According to spectrum estimation theory, the proposed scheme obtains the super-resolution estimations of the horizontal/vertical angle-of-arrivals (AoAs) and delays of sparse LoS paths with low overhead. Simulation results show that the proposed scheme can obtain a good channel estimation performance for multi-user wideband sparse scenario.