mmWave Channel Estimation via Approximate Message Passing with Side Information

Abstract

This work considers millimeter-wave channel estimation in a setting where parameters of the underlying mmWave channels are varying dynamically over time and there is a single drifting path. In this setting, channel estimates at time block $k$ can be used as side information (SI) when estimating the channel at block $k+1$. To estimate channel parameters, we employ an SI-aided (complex) approximate message passing algorithm and compare its performance to a benchmark based on orthogonal matching pursuit.