Measurement Matrix Optimization for Compressive Sensing-Based Method of Moments Based on Dynamic Extraction
Yang Liu, Zhonggen Wang, Lin Han, Wenyan Nie · 2025
To address the challenges in constructing the measurement matrix within the compressive sensing-based method of moments (CS-MoM) framework, this paper proposes a novel dynamic extraction method. The method evaluates the importance of each row in the impedance matrix using row norms and applies a dynamic extraction strategy, extracting rows with higher importance densely and those with lower importance sparsely, ensuring both row representativeness and the integrity of the global structure. Numerical analysis demonstrates that, compared to the uniform extraction-based CS-MoM method, the proposed method significantly enhances computational accuracy.