Global Optimization for Extinction Curve Reconstruction in Inverse Electromagnetic Scattering of Multiparticle Aggregates

Abstract

|Generalized Mie theory provides a theoretical solution to the extinction cross-section curve of an electromagnetic scattering event with a multiparticle aggregate, given the con(cid:12)gurational information of the constituent particles. However, deducing the con(cid:12)guration of the aggregate from the extinction cross-section curve is a nontrivial inverse problem that can be cast as a global optimization problem. To address this challenge, we propose a computational scheme that combines global optimization search algorithms with a calculator known as the Generalized Multiparticle Mie-solution. The scheme is tested using mock scattering cross-section curves based on randomly generated aggregate con(cid:12)gurations. The scheme successfully reproduces the scattering curve by minimizing the discrepancy between the two scattering curves. However, the ground-truth con(cid:12)guration is not reproduced, as initially expected. This is due to the inability of the global optimization algorithm scheme used in the present work to correctly locate the global minimum in the high-dimensional parameter space. Nonetheless, the partial success of the proposed scheme to reconstruct the mock curves provides an instructive experience for future attempts to solve the inverse electromagnetic scattering problem by (cid:12)ne-tuning the present approach.