What Is Computational Electromagnetics?

Computational electromagnetics is a discipline concerned with the numerical solution of Maxwell's equations to predict the behavior of electromagnetic fields in and around structures of engineering interest. Rather than solving Maxwell's equations analytically, which is feasible only for a small set of idealized geometries, computational methods discretize space and time or transform the continuous problem into a system of algebraic equations that a computer can solve to arbitrary precision. The field encompasses antenna design, electromagnetic compatibility, microwave circuit analysis, and the study of guided wave propagation.

The discipline emerged as a practical enterprise in the 1960s and 1970s, driven by the needs of antenna and radar engineering and enabled by the availability of digital computers. Two foundational contributions defined the modern era: Harrington's formulation of the method of moments for integral equation problems, and Yee's 1966 publication of the finite-difference time-domain lattice that bears his name. Subsequent decades brought finite element methods, hybrid solvers, and massively parallel implementations, expanding the scale of problems tractable by simulation.

Finite Difference Time Domain Method

The finite-difference time-domain (FDTD) method discretizes both space and time on a staggered Cartesian grid and advances the electric and magnetic field components in a leapfrog sequence governed by the curl equations of Maxwell. Introduced by Kane Yee in 1966, the method is a full-wave approach: a single time-domain simulation captures the system's response across a broad frequency range when the source is chosen as a short pulse. FDTD is well suited to problems involving dielectric and dispersive materials, biological tissue, and complex three-dimensional geometries. The Finite-Difference Time-Domain Method chapter in the Wiley-IEEE Press series provides a rigorous mathematical treatment of stability criteria, absorbing boundary conditions, and dispersive material modeling.

Method of Moments and Integral Equation Solvers

The method of moments, also known as the boundary element method in some communities, formulates an electromagnetic problem as an integral equation over the surfaces or volumes of conducting or dielectric objects and then discretizes that equation using a set of basis functions. Unlike volume-based methods, it requires meshing only surfaces, which reduces the number of unknowns for many antenna and scattering problems. The resulting matrix system is dense rather than sparse, so large problems demand fast solution algorithms such as the multilevel fast multipole method (MLFMM). Signal integrity analysis in high-speed circuit boards uses integral equation solvers to compute parasitic inductances and capacitances. Research on these methods appears regularly in the IEEE Transactions on Antennas and Propagation.

Transmission-Line Modeling and Monte Carlo Methods

The transmission-line modeling (TLM) method, developed by Johns and Beurle in 1971, models a three-dimensional electromagnetic field by an equivalent network of transmission lines on a Cartesian grid. It is time-domain like FDTD but derives from circuit theory rather than differential equations, making it natural for problems where network and field analyses must be combined. Monte Carlo methods enter computational electromagnetics in a different capacity: they are used to propagate uncertainty through electromagnetic simulations when material properties, geometric tolerances, or source conditions are known only probabilistically. The combination of field solvers with statistical sampling enables engineering tolerance analysis and reliable design for antenna arrays and electromagnetic shielding. Guided wave problems, including waveguide modes and surface plasmon propagation, are studied using all of these approaches depending on the geometry and frequency range. The IEEE Antennas and Propagation Society maintains a community of practice spanning all of these numerical methods and their experimental validation.

Applications

Computational electromagnetics has applications in a wide range of disciplines, including:

• Antenna and radar design, through full-wave simulation of radiation patterns and impedance
• Electromagnetic compatibility testing, predicting coupling between circuits and radiated emissions
• Biomedical engineering, modeling specific absorption rate (SAR) in tissue from RF devices
• Wireless communication, analyzing propagation in complex indoor and urban environments
• Photonics and nano-optics, simulating plasmonic structures and photonic crystal devices