Fast Construction of Measurement Matrix and Sensing Matrix With Dual Adaptive Cross Approximation

To accelerate the solution of current coefficients in the compressive sensing method of moments, a new strategy for constructing the measurement matrix is proposed, as well as the sensing matrix is optimized. First, the far-field group matrices in the impedance matrix are compressed using the adaptive cross approximation (ACA) algorithm, and the row indexes generated during the compression process are used as an extraction strategy rather than random or uniform extraction to construct the measurement matrix. Second, the ACA is used again, which approximates the far-field group matrices in the measurement matrix as the product of two low-rank matrices, resulting in a part-constructed sensing matrix. The dual ACA not only maintains computational accuracy, but also saves memory consumption, improving computational efficiency.