Improving the Performance of the Multilevel Matrix Decomposition Algorithm for 2.5-D Structures. Application to metamaterials

This paper presents an optimization of the multilevel matrix decomposition algorithm (MLMDA) that allows the efficient analysis of 2.5-D structures. This tool could be useful in the analysis of planar multilayer metamaterials (MM). Comparing its performance against the multilevel fast multiple algorithm (MLFMA) it was shown that MLMDA 2.5-D requires less computation time and provides better precision. Future work on the algorithm will be directed to take advantage of the periodicity present in MM