Perception-aided wideband millimeter-wave MIMO channel estimation: extended error modeling and robustness analysis

This paper proposes a sensing-assisted wideband millimeter-wave (mmWave) MIMO channel estimation algorithm combining Simultaneous Weighted Orthogonal Matching Pursuit (SWOMP) and Sparse Bayesian Learning (SBL). To better characterize sensing uncertainty, the study extends conventional Gaussian error modeling by introducing uniform and truncated Gaussian distributions, reflecting realistic radar sensing characteristics such as bounded and quantized deviations. The proposed two-stage SWOMP–SBL framework achieves high-accuracy channel reconstruction and strong robustness across various error conditions. Simulation results show that the algorithm maintains nearly identical NMSE performance under Gaussian and truncated Gaussian errors, with only minor degradation under uniform noise. Moreover, the framework effectively reduces pilot overhead and computational complexity while ensuring convergence stability. These results demonstrate that the proposed approach provides a practical and efficient solution for robust channel estimation in next-generation Integrated Sensing and Communication (ISAC) systems for 6G mmWave/THz networks.