Estimation of Frequency-Selective Channels for Millimeter Wave MIMO Systems based on Compressive Sensing

Channel estimation is one of the key issues in millimeter wave (mmWave) multiple input and multiple output (MIMO) systems. This paper considers frequency-selective mmWave channels and proposes a channel estimation algorithm based on compressive sensing (CS) theory. By exploiting the joint sparse structure between channels at different subcarriers, the proposed algorithm can estimate the channels at all subcarriers simultaneously and the performance of channel estimation can be improved effectively. Simulation results demonstrate that the proposed algorithm performs better than traditional CS-based algorithm in mmWave channel estimation.