Reconstruction of the refractive index from experimental backscattering data using a globally convergent inverse method

TL;DR

This paper evaluates a globally convergent inverse algorithm for reconstructing the refractive index of inhomogeneous media from experimental backscattering data, addressing data mismatch issues and demonstrating effective results on real targets.

Contribution

It applies and analyzes a globally convergent inverse method to real experimental data, including data pre-processing to improve reconstruction accuracy.

Findings

Successful reconstruction of high-contrast targets.

Effective data pre-processing for experimental data.

Good performance on both blind and non-blind targets.

Abstract

The problem to be studied in this work is within the context of coefficient identification problems for the wave equation. More precisely, we consider the problem of reconstruction of the refractive index (or equivalently, the dielectric constant) of an inhomogeneous medium using one backscattering boundary measurement. The goal of this paper is to analyze the performance of a globally convergent algorithm of Beilina and Klibanov on experimental data acquired in the Microwave Laboratory at University of North Carolina at Charlotte. The main challenge working with experimental data is the the huge misfit between these data and computationally simulated data. We present data pre-processing steps to make the former somehow look similar to the latter. Results of both non-blind and blind targets are shown indicating good reconstructions even for high contrasts between the targets and the background medium.