Stabilized and Fast Method for Compressive-Sensing-Based Method of Moments
Yalan Gao, Muhammad Firdaus Akbar, Ghassan Nihad Jawad
Abstract
To improve the stability and time performance of the compressive-sensing-based method of moments (CS–MoM) for bistatic electromagnetic scattering problems of three-dimensional objects, a novel method is proposed in this letter. First, the measurement matrix constructed by random extraction in the conventional CS–MoM is theoretically and numerically proven to be irrational and causes unstable results. And a deterministic measurement matrix constructed with fixed-step extraction strategy is proposed to eliminate the instability. Second, based on the modal significance and the algorithmic structure of generalized orthogonal matching pursuit (gOMP), the theory of constructing sparse basis using characteristic modes is reformulated, and the gOMP is simplified to least squares for solving the final constructed overdetermined system. Compared with the conventional CS–MoM, the suggested method can provide more stable results and less time cost. The results of various numerical simulations attest to the accuracy and effectiveness of the suggested method.