The Multilevel Fast Multipole Algorithm (MLFMA) for Solving Large-Scale Computational Electromagnetics Problems

Preface

This book is about a sophisticated algorithm, namely, the multilevel fast multipole algorithm (MLFMA), which has been developed and improved continuously in the last two decades for the fast and accurate solution of real-life electromagnetics problems. When it was first proposed, MLFMA enabled accurate solutions of problems relating to metallic objects that cannot be solved directly on computers. Since then, the algorithm has been extended to more complicated problems involving metallic, dielectric, and composite objects, has been improved for realistic solutions with unprecedented levels of accuracy, and has been parallelized to handle discretizations involving millions of unknowns. Recently, MLFMA has been the core algorithm to solve the largest dense matrix equations involving more than 1 billion unknowns.

Due to high academic and industrial impacts, computational simulations of electromagnetic phenomena have attracted the interest of many researchers. The literature includes excellent studies involving huge efforts to develop novel solvers for more efficient and accurate simulations. Along this direction, MLFMA has been one of the major algorithms that have passed the time test. Since it was proposed, this algorithm has proved its robustness, efficiency, and accuracy in many simulations involving diverse components and parameters. Today, most of the leading research centers now have their own implementations of MLFMA, for sequential and/or parallel platforms, among ...