Hierarchical Channel Estimation for Near-Field Spatial Non-Stationary Channels: A Pre-Selection and Multi-Level Dynamic Threshold Strategy

This paper proposes a robust hierarchical channel estimation algorithm for massive multiple-input multiple-output (MIMO) systems, specifically addressing the challenges posed by near-field spatial non-stationary channels. In near-field communication scenarios, the channel characteristics exhibit significant spatial variations due to the proximity between the transmitter and receiver, resulting in non-uniform sparsity patterns that hinder traditional estimation methods. To enhance estimation efficiency and accuracy, we propose an approach that integrates a pre-selection strategy with a multi-level dynamic thresholding mechanism. The proposed algorithm operates in two stages. In the first stage, a pre-selection process effectively reduces the number of candidate atoms, improving the computational efficiency of sparse adaptive estimation. In the second stage, a dynamic multi-level thresholding scheme is introduced, where the noise reconstruction parameter is adaptively adjusted based on the instantaneous signal-to-noise ratio (SNR), ensuring robustness across varying SNR conditions. Simulation results demonstrate that the proposed method outperforms existing algorithms in terms of mean square error (MSE) and reconstruction success probability while maintaining computational complexity comparable to conventional approaches. Given its superior performance and efficiency, the proposed algorithm is well-suited for deployment in near-field MIMO systems, making it a promising solution for next-generation wireless networks.