Norm-1 Regularized Consensus-based ADMM for Imaging with a Compressive Antenna

TL;DR

This paper introduces a norm-one-regularized, consensus-based ADMM algorithm for imaging with a compressive antenna, achieving faster convergence and enabling quasi-real-time imaging of dielectric and metallic targets with limited data.

Contribution

It proposes a novel ADMM-based imaging algorithm that outperforms existing methods in computational efficiency and leverages a compressive reflector antenna for high-capacity sensing.

Findings

Outperforms Nesterov-based algorithms in computational cost

Enables quasi-real-time imaging with compressive antennas

Effective for imaging composite dielectric and metallic targets

Abstract

This paper presents a novel norm-one-regularized, consensus-based imaging algorithm, based on the Alternating Direction Method of Multipliers (ADMM). This algorithm is capable of imaging composite dielectric and metallic targets by using limited amount of data. The distributed capabilities of the ADMM accelerates the convergence of the imaging. Recently, a Compressive Reflector Antenna (CRA) has been proposed as a way to provide high-sensing-capacity with a minimum cost and complexity in the hardware architecture. The ADMM algorithm applied to the imaging capabilities of the Compressive Antenna (CA) outperforms current state of the art iterative reconstruction algorithms, such as Nesterov-based methods, in terms of computational cost; and it ultimately enables the use of a CA in quasi-real-time, compressive sensing imaging applications.