Fast field back-propagation using a multilevel non-uniform grid algorithm in oblate spheroidal coordinates

Abstract

The Rayleigh-Sommerfeld (RS) back-propagation method is accelerated by a modified version of the multilevel nonuniform grid algorithm (MLNG). The algorithm utilizes oblate spheroidal grids to be effective in terms of the number of sampling points that are required for correct reconstruction of the fields measured on quasi-planar surfaces. The method is demonstrated on the example of a parabolic reflector with a surface distortion. A good agreement between the results of the direct RS integration and the proposed method is achieved.

2D and 3D sampling schemes for fast multilevel field back-propagation algorithm

“…For radiating surfaces that are very large compared to the wavelength, this computational bottleneck renders this approach unattractive. Significant reduction of the CC down to O log is achieved using a modified version of the multilevel non-uniform grid (MLNG) [2], [3]. The MLNG technique is based on a hierarchical divide-andconquer strategy.…”