Phase Retrieval and System Identification in Dynamical Sampling via Prony's Method

TL;DR

This paper introduces a method combining phase retrieval and system identification in dynamical sampling using Prony's method, providing constructive guarantees and error bounds applicable in finite and infinite-dimensional spaces.

Contribution

It develops a novel approach that jointly recovers signals and unknown system components in dynamical sampling by leveraging Prony's method, with proven guarantees.

Findings

Recovery guarantees for signals and systems using Prony's method.

Constructive analytic recovery procedures applicable in finite and infinite dimensions.

Error bounds for approximate Prony method in complex exponential sum recovery.

Abstract

Phase retrieval in dynamical sampling is a novel research direction, where an unknown signal has to be recovered from the phaseless measurements with respect to a dynamical frame, i.e. a sequence of sampling vectors constructed by the repeated action of an operator. The loss of the phase here turns the well-posed dynamical sampling into a severe ill-posed inverse problem. In the existing literature, the involved operator is usually completely known. In this paper, we combine phase retrieval in dynamical sampling with the identification of the system. For instance, if the dynamical frame is based on a repeated convolution, then we want to recover the unknown convolution kernel in advance. Using Prony's method, we establish several recovery guarantees for signal and system, whose proofs are constructive and yield analytic recovery methods. The required assumptions are satisfied by almost all signals, operators, and sampling vectors. Moreover, these guarantees not only hold for the finite-dimensional setting but also carry over to infinite-dimensional spaces. Studying the sensitivity of the analytic recovery procedures, we also establish error bounds for the applied approximate Prony method with respect to complex exponential sums.