Compressed Channel Estimation for Point-to-Point Millimeter-Wave Communications

In this paper, a compressed sensing based channel estimation is proposed for point-to-point where the single-carrier waveform is used over the millimeter-wave spectrum. The key idea of the proposed scheme referred to as randomly sampled channel impulse recovery (RS-CIR), is to estimate channel impulse response based on randomly selected time-domain pilot samples transmitted in a localized portion of the ultra-wide millimeter-wave bandwidth. As a result of RS-CIR, channel information of full bandwidth can be acquired from compressed sensing without having to take the measurement on the entire bandwidth. From the numerical evaluations, it is observed that the proposed scheme outperforms conventional channel estimation schemes under the situation of sharp beamforming or line-of-sight condition in terms of estimation accuracy, achieving significant pilot overhead reduction.