Current browse context:
stat.CO
Change to browse by:
References & Citations
Statistics > Computation
Title: Convolutional Phase Retrieval via Gradient Descent
(Submitted on 3 Dec 2017 (v1), last revised 6 Oct 2019 (this version, v3))
Abstract: We study the convolutional phase retrieval problem, of recovering an unknown signal $\mathbf x \in \mathbb C^n $ from $m$ measurements consisting of the magnitude of its cyclic convolution with a given kernel $\mathbf a \in \mathbb C^m $. This model is motivated by applications such as channel estimation, optics, and underwater acoustic communication, where the signal of interest is acted on by a given channel/filter, and phase information is difficult or impossible to acquire. We show that when $\mathbf a$ is random and the number of observations $m$ is sufficiently large, with high probability $\mathbf x$ can be efficiently recovered up to a global phase shift using a combination of spectral initialization and generalized gradient descent. The main challenge is coping with dependencies in the measurement operator. We overcome this challenge by using ideas from decoupling theory, suprema of chaos processes and the restricted isometry property of random circulant matrices, and recent analysis of alternating minimization methods.
Submission history
From: Qing Qu [view email][v1] Sun, 3 Dec 2017 06:04:25 GMT (2573kb)
[v2] Wed, 28 Aug 2019 03:30:06 GMT (5643kb,D)
[v3] Sun, 6 Oct 2019 02:55:12 GMT (5604kb,D)
Link back to: arXiv, form interface, contact.