Rao-Blackwellised Interacting Markov Chain Monte Carlo for Electromagnetic Scattering Inversion

TL;DR

This paper presents a Bayesian approach using advanced Markov Chain Monte Carlo methods to estimate electromagnetic material properties from scattering data, leveraging high-performance computing for large-scale inverse problems.

Contribution

It introduces a Rao-Blackwellised interacting MCMC method tailored for electromagnetic scattering inversion, improving estimation accuracy and uncertainty quantification.

Findings

Efficiently estimates material properties from large scattering datasets.

Demonstrates the effectiveness of SMC methods in high-dimensional inverse problems.

Provides uncertainty quantification for electromagnetic property estimates.

Abstract

The following electromagnetism (EM) inverse problem is addressed. It consists in estimating local radioelectric properties of materials recovering an object from the global EM scattering measurement, 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 (HPC) 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 dynamic stochastic process, evolving in function of the frequency, it is shown how advanced Markov Chain Monte Carlo methods, called Sequential Monte Carlo (SMC) or interacting particles, can provide estimations of the EM properties of each material, and their associated uncertainties.