Frequency-domain Compressive Channel Estimation for Frequency-Selective Hybrid mmWave MIMO Systems

摘要

Channel estimation is useful in millimeter wave (mmWave) MIMO communicationsystems. Channel state information allows optimized designs of precoders andcombiners under different metrics such as mutual information orsignal-to-interference-noise (SINR) ratio. At mmWave, MIMO precoders andcombiners are usually hybrid, since this architecture provides a means totrade-off power consumption and achievable rate. Channel estimation ischallenging when using these architectures, however, since there is no directaccess to the outputs of the different antenna elements in the array. The MIMOchannel can only be observed through the analog combining network, which actsas a compression stage of the received signal. Most of prior work on channelestimation for hybrid architectures assumes a frequency-flat mmWave channelmodel. In this paper, we consider a frequency-selective mmWave channel andpropose compressed-sensing-based strategies to estimate the channel in thefrequency domain. We evaluate different algorithms and compute their complexityto expose trade-offs in complexity-overhead-performance as compared to those ofprevious approaches.