Title: Blind Reconstruction of Multilayered Tissue Profiles with UWB Radar Under Bayesian Setting

Abstract: In this paper, we investigate the problem of inverse electromagnetic scattering to recover multilayer human tissue profiles using ultrawideband radar systems in a Bayesian setting. We study the recovery problem in a blind setting, in which we simultaneously estimate both the dielectric/geometric properties of the one-dimensional target tissue profile and the transmitted radar waveform. To perform Bayesian parameter estimation, we propose a hybrid and adaptive Markov Chain Monte Carlo method, which combines the Slice sampling and Hamiltonian Monte Carlo approaches. The introduced sampling mechanism also incorporates the Parallel Tempering approach to escape from the local optimal regions of the complex posterior distribution. We provide empirical support through various numerical simulations for the achieved enhanced sampling efficiency compared to conventional sampling schemes. To investigate the recovery performance, we work on synthetic measurements simulating actual radar returns from multilayer tissue profiles. We derive theoretical bounds for the best achievable estimation performance in terms of normalized root mean square error and provide a comparison with the performance of our estimator.