Advanced Interacting Sequential Monte Carlo Sampling for Inverse Scattering

TL;DR

This paper presents an advanced sequential Monte Carlo sampling method for solving large-scale, ill-posed inverse electromagnetic scattering problems by estimating material properties from scattering data using high-performance computing and Bayesian inference.

Contribution

It introduces an improved interacting SMC approach tailored for inverse EM problems, leveraging problem structure and high-performance computing for efficient Bayesian estimation.

Findings

Effective estimation of radioelectric properties from scattering data.

Utilization of high-performance computing for large-scale inverse problems.

Improved accuracy over traditional methods in local EM property estimation.

Abstract

The following electromagnetism (EM) inverse problem is addressed. It consists in estimating local radioelectric properties of materials recovering an object from global EM scattering measurements, at various incidences and wave frequencies. This large scale ill-posed inverse problem is explored by an intensive exploitation of an efficient 2D Maxwell solver, distributed on high performance computing machines. Applied to a large training data set, a statistical analysis reduces the problem to a simpler probabilistic metamodel, on which Bayesian inference can be performed. Considering the radioelectric properties as a hidden dynamic stochastic process, that evolves in function of the frequency, it is shown how advanced Markov Chain Monte Carlo methods, called Sequential Monte Carlo (SMC) or interacting particles, can take benefit of the structure and provide local EM property estimates.