Efficient Low-Overhead Channel Estimation for 5G Lens Based Millimeter-Wave Massive MIMO Systems

Beamspace MIMO performs beam-selection which can substantially reduce the number of power-consuming radio frequency (RF) chains without perceptible performance deterioration. However, for capacityapproaching performance, accurate information of the beamspace-channel of large-size is required for beamselection, which is contesting in case of little number of RF-chains. To overcome such problem, I proposed an efficient support-detection (SD) algorithm for channel-estimation with low pilot-overhead and short number of RF chains. The key idea of SD-algorithm is to divide the whole issue of beamspace channel-estimation into a series of sub-issues, where each of them considers only one sparse channel-component. The support of each channel component is detected reliably by deploying the sparse structure attributes of the beamspace-channel. The effect of this channel-component is eliminated from the whole channel-estimation issue. Thus, the sparse beamspace-channel can be estimated with low pilot-overhead. Simulation Results shows that the proposed schemes perform much better than the conventional compressed-sensing (CS) schemes.