Sparse Channel Estimation for Stacked Intelligent Metasurface-Assisted mmWave Communications

Although massive multiple-input multiple-output (MIMO) and millimeter-wave (mmWave) technologies show great promise in enhancing the spectral and energy efficiency of wireless networks, their practical implementation is constrained by energy consumption and hardware costs. Particularly, mmWave communication faces challenges due to high atmospheric attenuation, requiring expensive and energy-intensive transceivers. Additionally, poor scattering conditions may compromise the spatial multiplexing gains of massive MIMO systems. Recently, the emerging technology of stacked intelligent metasurfaces (SIM) has gained prominence due to its low power consumption and reduced number of radio frequency (RF) chains, making it a compelling candidate for mmWave communication. Specifically, SIM resembles an artificial neural network (ANN), comprising multiple layers of metasurfaces with numerous tunable passive meta-atoms, allowing flexible manipulation of electromagnetic waves for beamforming. In this paper, we investigate the channel estimation problem in SIM-assisted mmWave communication systems. Given that the number of meta-atoms far exceeds the number of base station (BS) antennas, obtaining channel state information presents a challenge. We introduce a mutually orthogonal pilot strategy to mitigate multi-user interference and apply a hard threshold orthogonal matching pursuit (HT-OMP) algorithm for sparse channel estimation. Numerical results validate the effectiveness of the proposed approach.