Atomic norm denoising with applications to line spectral estimation

TL;DR

This paper introduces a new atomic norm-based method for line spectral estimation that provides theoretical error guarantees, scalable algorithms, and outperforms classical methods in noisy conditions.

Contribution

It develops a convex optimization framework for spectral estimation using atomic norms, with polynomial-time solvability and scalable approximations.

Findings

SDP-based approach outperforms classical methods in MSE

L1-regularized least squares approximates SDP with similar error

Method provides theoretical guarantees without knowing model order

Abstract

Motivated by recent work on atomic norms in inverse problems, we propose a new approach to line spectral estimation that provides theoretical guarantees for the mean-squared-error (MSE) performance in the presence of noise and without knowledge of the model order. We propose an abstract theory of denoising with atomic norms and specialize this theory to provide a convex optimization problem for estimating the frequencies and phases of a mixture of complex exponentials. We show that the associated convex optimization problem can be solved in polynomial time via semidefinite programming (SDP). We also show that the SDP can be approximated by an l1-regularized least-squares problem that achieves nearly the same error rate as the SDP but can scale to much larger problems. We compare both SDP and l1-based approaches with classical line spectral analysis methods and demonstrate that the SDP outperforms the l1 optimization which outperforms MUSIC, Cadzow's, and Matrix Pencil approaches in terms of MSE over a wide range of signal-to-noise ratios.