Prony-Based Super-Resolution Phase Retrieval of Sparse, Multivariate Signals

TL;DR

This paper introduces a Prony-based super-resolution method for phase retrieval of sparse multivariate signals, utilizing adaptive sampling along specific lines to achieve unique recovery with fewer measurements.

Contribution

It provides a novel recovery guarantee for multivariate sparse signals using adaptive Fourier sampling, extending Prony's method to phase retrieval.

Findings

Recovery guaranteed up to global ambiguities

Sampling complexity scales quadratically with sparsity

Method effective for bivariate signals with generic directions

Abstract

Phase retrieval consists in the recovery of an unknown signal from phaseless measurements of its usually complex-valued Fourier transform. Without further assumptions, this problem is notorious to be severe ill posed such that the recovery of the true signal is nearly impossible. In certain applications like crystallography, speckle imaging in astronomy, or blind channel estimation in communications, the unknown signal has a specific, sparse structure. In this paper, we exploit these sparse structure to recover the unknown signal uniquely up to inevitable ambiguities as global phase shifts, transitions, and conjugated reflections. Although using a constructive proof essentially based on Prony's method, our focus lies on the derivation of a recovery guarantee for multivariate signals using an adaptive sampling scheme. Instead of sampling the entire multivariate Fourier intensity, we only employ Fourier samples along certain adaptively chosen lines. For bivariate signals, an analogous result can be established for samples in generic directions. The number of samples here scales quadratically to the sparsity level of the unknown signal.