What Are Integral Equations?

Integral equations are equations in which an unknown function appears under an integral sign, making them distinct from differential equations, where the unknown appears in derivative form. The solution to an integral equation is a function, and that function typically appears both inside and outside the integral. Integral equations arise naturally in physics, engineering, and applied mathematics wherever a quantity at one point depends on the weighted influence of the same quantity distributed across a region or boundary. They are fundamental tools in electromagnetic scattering theory, potential theory, elasticity, heat transfer, and signal processing.

The classical taxonomy of integral equations divides them into Fredholm equations, where the limits of integration are fixed, and Volterra equations, where one integration limit is a variable. Each class is further divided into first-kind equations, where the unknown appears only inside the integral, and second-kind equations, where the unknown appears both inside and outside. Integrodifferential equations, which combine integral and differential operators, arise in viscoelasticity and population dynamics. The mathematical foundation draws on functional analysis, particularly the theory of compact operators and Hilbert spaces.

Fredholm and Volterra Equations

A Fredholm equation of the second kind has the form: the unknown function equals a known source term plus a kernel function integrated against the unknown over a fixed domain. The kernel function encodes the physical or mathematical relationship governing the problem. When the kernel is symmetric and the operator it defines is compact, spectral theory guarantees the existence and structure of solutions. Volterra equations of the second kind share a similar structure but with a variable upper limit of integration, which gives them a causal character suited to time-domain problems such as hereditary mechanics. Numerical solution of both types typically uses quadrature discretization, converting the integral into a matrix-vector product, or Galerkin and collocation methods that expand the unknown in a basis of trial functions.

Boundary-Element Methods

The boundary-element method (BEM) reformulates a partial differential equation problem in a volume domain as an integral equation on the bounding surface, reducing the spatial dimension of the problem by one. For Laplace, Helmholtz, and Maxwell equations, this reformulation is possible because fundamental solutions (Green's functions) are known analytically. The boundary element method applied to solving Fredholm and Volterra integral equations converts the original equation into a system of algebraic equations by discretizing the boundary into elements and applying Galerkin or collocation projection. BEM is particularly effective for unbounded domains, such as exterior electromagnetic or acoustic scattering, where volumetric discretization methods would require an artificial boundary truncation. Research on Galerkin boundary element methods for electromagnetic scattering provides rigorous convergence estimates for the boundary element solution of the electric field integral equation arising from time-harmonic Maxwell's equations.

Method of Moments

The method of moments (MoM) is the dominant integral equation technique in computational electromagnetics, particularly for antenna analysis and scattering from conducting bodies. MoM expands the unknown surface current in a set of basis functions (commonly RWG functions on triangular meshes), substitutes the expansion into the electric field integral equation, and tests the residual against a set of weighting functions to produce a dense linear system. The IEEE Xplore reference on antenna modeling using integral equations and the method of moments is a standard treatment of this methodology for wire and surface antenna problems. MoM matrices are dense, requiring O(N²) memory, but fast algorithms such as the multilevel fast multipole method reduce this to O(N log N).

Applications

Integral equations have applications across a range of technical and scientific fields, including:

• Electromagnetic scattering and antenna analysis in wireless communications and radar
• Inverse problems in medical imaging, including electrical impedance tomography
• Geophysical survey data inversion for subsurface characterization
• Deconvolution in signal processing and optical system analysis
• Elasticity and fracture mechanics in structural engineering
• Acoustic radiation and noise prediction in aeronautics