Sparse phase retrieval via Phaseliftoff

DOI10.1109/TSP.2021.3067164MaRDI QIDQ6347447arXivFDO

Publication date 20 August 2020

Signal theory (characterization, reconstruction, filtering, etc.) (94A12)

Abstract: The aim of sparse phase retrieval is to recover a

k

-sparse signal

m a t h b f x_{0} i n m a t h b b C^{d}

from quadratic measurements

| l a n g l e m a t h b f a_{i}, m a t h b f x_{0} a n g l e |^{2}

where

m a t h b f a_{i} i n m a t h b b C^{d}, i = 1, l d o t s, m

. Noting

| l a n g l e m a t h b f a_{i}, m a t h b f x_{0} a n g l e |^{2} = e x t T r (A_{i} X_{0})

with

A_{i} = m a t h b f a_{i} m a t h b f a_{i}^{*} i n m a t h b b C^{d i m e s d}, X_{0} = m a t h b f x_{0} m a t h b f x_{0}^{*} i n m a t h b b C^{d i m e s d}

, one can recast sparse phase retrieval as a problem of recovering a rank-one sparse matrix from linear measurements. Yin and Xin introduced PhaseLiftOff which presents a proxy of rank-one condition via the difference of trace and Frobenius norm. By adding sparsity penalty to PhaseLiftOff, in this paper, we present a novel model to recover sparse signals from quadratic measurements. Theoretical analysis shows that the solution to our model provides the stable recovery of

m a t h b f x_{0}

under almost optimal sampling complexity

m = O (k l o g (d / k))

. The computation of our model is carried out by the difference of convex function algorithm (DCA). Numerical experiments demonstrate that our algorithm outperforms other state-of-the-art algorithms used for solving sparse phase retrieval.

This page was built for publication: Sparse phase retrieval via Phaseliftoff

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6347447)