Memory-Efficient Modeling of Reverberation Chambers Using Hybrid Recursive Update Discrete Singular Convolution-Method of Moments

Abstract

A hybrid method is proposed for memory-efficient analysis of reverberation chambers (RCs). In the hybrid method, the cavity is modeled by the recursive update discrete singular convolution (RUDSC) method, and antennas and stirrers inside the cavity are simulated using the flexible method of moments (MoM). In order to solve DSC and MoM unknowns separately, a layer-based elimination algorithm is utilized to eliminate the DSC unknowns. The MoM unknowns are then solved by a direct solver. Once the solution for the MoM model is obtained, the original RC is equivalent to a cavity excited by known current sources. The equivalent problem is finally solved using the RUDSC method. Taking advantage of the layer-based elimination algorithm and the recursive update technique, the memory requirement of the new hybrid method is much smaller than that of using a direct solver. Numerical simulations are presented to show the efficacy of the proposed method. It is shown that the the proposed method substantially reduces the memory cost of RC modeling, which extends RC analysis to higher frequencies.