Channel estimation for IRS-Aided mmWave massive MIMO systems based on variational Bayesian method with Laplacian prior

Intelligent Reflecting Surface (IRS) aided mmWave Massive Multiple-Input Multiple Output (MIMO) system has been considered as a key enabler for future 6G systems. The excellent performance of such systems highly relies on accurate channel state information (CSI) which, however, is very challenging to obtain. To solve this problem, in this work we propose a channel estimation scheme based on variational Bayesian compressive sensing with Laplace prior (VBCS-Laplace). The channel estimation problem is reconstructed as in a sparse form and a sensing matrix is designed based on angular domain quantization. Different from existing works, we propose to employ Laplace prior to exploit the inherent space sparsity in mmWave channel. An iterative algorithm is proposed to update the hyper-parameters in the Bayesian model. To verify the performance, extensive simulations are carried out and numerical results show that the proposed VBCS-Laplace significantly outperforms the state-of-the-art counterparts with a slightly increase in computational complexity.