Moment Method

What Is the Moment Method?

The moment method is a numerical technique for solving integral equations that arise in electromagnetics, acoustics, and other physics domains by converting them into a finite system of linear algebraic equations. Originally developed in a general mathematical context, the method was introduced to computational electromagnetics by Roger Harrington in his 1968 monograph, after which it became the standard approach for analyzing antennas, scattering bodies, and microwave circuits when exact analytical solutions are unavailable. The technique belongs to the broader class of weighted-residual methods and is closely related to Galerkin procedures, though the choice of basis and testing functions can vary. Its principal strength is that it operates directly on the electromagnetic boundary, rather than discretizing a volume, which reduces the dimensionality of the problem and makes it efficient for open-region problems such as radiating structures in free space.

Mathematical Formulation and Basis Functions

The moment method begins by expressing the unknown quantity, typically a surface current density on a conductor or an aperture field, in terms of a set of basis functions multiplied by unknown coefficients. Commonly used bases include pulse functions for their simplicity, triangular subdomain functions for piecewise linearity, and the Rao-Wilton-Glisson (RWG) triangular elements for modeling curved metallic surfaces with high geometric fidelity. The original integral equation is then enforced in a weighted-residual sense by multiplying by a set of testing functions and integrating over the domain. When the testing functions match the basis functions, the result is a Galerkin formulation, which typically yields better convergence than point-matching alternatives. The process produces a dense, complex-valued matrix equation, Z·I = V, where Z is the impedance matrix encoding mutual interactions among all basis elements, I is the vector of unknown coefficients, and V represents the excitation. The IEEE Xplore chapter on the Method of Moments in Applied Frequency-Domain Electromagnetics provides a rigorous derivation of these steps.

Computational Implementation

Solving the dense linear system resulting from a moment method formulation requires careful attention to memory and computation time. A problem with N basis functions requires storing and inverting an N-by-N complex matrix, so memory scales as N-squared and direct solution time scales as N-cubed. For large problems, iterative solvers such as the conjugate gradient method or GMRES are used in place of direct Gaussian elimination. The multilevel fast multipole algorithm (MLFMA) reduces the matrix-vector product time from N-squared to approximately N·log(N), enabling the moment method to scale to problems with millions of unknowns. Software implementations such as MININEC, a method-of-moments program for thin-wire antennas, demonstrated early on how the technique could be packaged for practical engineering use, as documented in IEEE work on the evolution of MININEC for antenna analysis. Parallelization across multicore processors and GPU acceleration have further extended the practical limit of solvable problem sizes.

Convergence and Accuracy

The accuracy of a moment method solution depends on both the density of the mesh and the order of the basis functions. Higher-order bases reduce the number of unknowns needed for a given accuracy but complicate integration over curved surfaces. A useful rule of thumb is ten unknowns per wavelength for first-order basis functions on smooth surfaces. Singularity extraction techniques handle the near-singular behavior of Green's functions when source and observation points coincide, a step critical to matrix accuracy. Validation against analytical solutions, such as Mie scattering from a sphere, is standard practice for confirming that an implementation is converging correctly. An efficient integral equation method for full-wave analysis published on arXiv illustrates recent efforts to extend moment method accuracy to broadband electromagnetic problems.

Applications

The moment method has applications in a range of fields, including: