Parametric Diffraction-Based Object Sensing: Modeling, Estimation, and Fundamental Limits

This paper proposes a rigorous framework for sensing of environmental objects using diffraction mechanisms prevalent at wireless communication frequencies.

Specifically, we develop a physics-consistent parameterized diffraction channel model, derive maximum likelihood (ML) approaches for estimating the blockage shape, range, and source directions of arrival (DoAs), and quantify fundamental performance limits via the Cramér--Rao bound (CRB).

In our physics-based modeling, we integrate various approximations for the wave propagation (far-field, paraxial Fresnel, and exact near-field regimes), enabling a wide range of applicability.

The underlying model is frequency-agnostic, and we derive Fresnel-number scaling laws that map the diffraction pattern, and hence the estimation problem, across carrier frequency, object size, and range.

We quantify the maximum likelihood estimation performance and its relationship to the CRB, and we study the impact of the modeling approximations developed in this work.

Numerical results demonstrate that ML estimators closely approach the CRB at moderate to high signal-to-noise ratio (SNR), and highlight the utility of diffraction-based modeling for high-fidelity blockage characterization.