Irregular Array Manifold Aided Channel Estimation in Massive MIMO Communications

Angle information aided channel estimation has recently attracted extensive attention since its exploitation of the underlying angular channel structure has brought significant performance improvement. Nevertheless, current research works are mostly tailored to one dimensional (1D) or two dimensional (2D) antenna arrays. Being heavily dependent on the shift-invariant structure of the arrays, these existing results cannot be easily extended to irregular three dimensional (3D) arrays, which however are indispensable for base stations with large numbers of antennas but limited space. To extend the angle aided channel estimation to the regime of 3D irregular arrays, the channel estimation problem for an irregularly shaped array is linked to an array manifold constrained complex-valued tensor decomposition problem under missing data, and an efficient two-stage channel estimation algorithm that can automatically estimate the channel path number is further proposed. Simulation results are presented to demonstrate the excellent performance of the proposed algorithm in channel estimation, including path number enumeration.