Dynamic Validation SWOMP for Channel Estimation Using Training-Based mmWave MIMO Channels

This letter proposes a dynamic validation-based simultaneous weighted orthogonal matching pursuit (SWOMP) algorithm for accurate channel estimation using training-based millimeter-wave (mmWave) multiple input multiple output (MIMO) channels. The proposed method enhances traditional SWOMP by introducing a dynamic path validation mechanism, which adaptively selects the most significant paths based on their contribution to minimizing the residual norm. Leveraging training channel data, the algorithm dynamically optimizes the sparsity level, ensuring robust performance in frequency-selective and varying channel conditions. Simulation results show that the proposed algorithm outperforms existing SWOMP variants and other compressed sensing-based methods in estimation accuracy and robustness across diverse SNR levels.