Single-sided adaptive estimation of multi-path millimeter wave channels

Millimeter wave (mmWave) cellular systems will enable ultra high data rates by communicating over the large bandwidth available in mmWave frequencies. To overcome the channel propagation characteristics in this frequency band, large antenna arrays need to be deployed at both the base station and mobile users. While these large arrays provide sufficient beamforming gains to meet the required link margins, they make it challenging to estimate the mmWave channel. In this paper, we propose a mmWave channel estimation algorithm that exploits the sparse nature of the channel and leverages tools from adaptive compressed sensing to efficiently estimate the channel with a small training overhead. The proposed algorithm considers practical hardware constraints on the training beamforming design, and does not require the availability of a feedback channel between the base station and the mobile user. Simulation results indicate that comparable precoding gains can be achieved by the proposed channel estimation algorithm relative to the case when perfect channel knowledge exists.