Super-resolution channel estimation for mmwave massive MIMO

Millimeter-wave massive MIMO with hybrid precoding can significantly reduce the number of required radio frequency (RF) chains. However, due to the limited number of RF chains, the requirement of large pilot overhead will be used to estimate the high-dimensional mmWave massive MIMO channel. To solve this problem, we propose a super-resolution channel estimation scheme based on modified two- dimensional (2D) unitary ESPRIT algorithm in this paper, which can jointly estimate the continuously distributed angle of arrivals/departures (AoAs/AoDs) with high accuracy. Specifically, we first estimate a low-dimensional effective channel by designing the uplink training signals. Then, since the low- dimensional effective channel has the same shift- invariance of array response as the highdimensional mmWave MIMO channel to be estimated, we jointly estimate the super-resolution estimates of AoAs and AoDs by exploiting the modified 2D unitary ESPRIT algorithm. Furthermore, we can obtain the associated path gains based on the least squares (LS) criterion. Finally, the high-dimensional mmWave MIMO channel will be reconstructed according to the obtained parameters. Simulation results verify that the proposed scheme is more accuracy than conventional schemes, even with a much lower pilot overhead.