Leveraging Diversity and Sparsity in Blind Deconvolution

DOI10.1109/TIT.2017.2788444MaRDI QIDQ5375559zbMATH OpenWikidataFDO

Authors

Publication date 14 September 2018

Published in IEEE Transactions on Information Theory (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1610.06098

Mathematics Subject Classification ID

Communication theory (94A05) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Signal theory (characterization, reconstruction, filtering, etc.) (94A12)

Abstract: This paper considers recovering $L$-dimensional vectors $$, and $$ from their circular convolutions $$. The vector $$ is assumed to be $S$-sparse in a known basis that is spread out in the Fourier domain, and each input $$ is a member of a known $K$-dimensional random subspace. We prove that whenever $K+Slo{g}^{2}SlesssimL/lo{g}^4(LN)$, the problem can be solved effectively by using only the nuclear-norm minimization as the convex relaxation, as long as the inputs are sufficiently diverse and obey $Ngtrsimlo{g}^2(LN)$. By "diverse inputs", we mean that the $$'s belong to different, generic subspaces. To our knowledge, this is the first theoretical result on blind deconvolution where the subspace to which $$ belongs is not fixed, but needs to be determined. We discuss the result in the context of multipath channel estimation in wireless communications. Both the fading coefficients, and the delays in the channel impulse response $$ are unknown. The encoder codes the $K$-dimensional message vectors randomly and then transmits coded messages $$'s over a fixed channel one after the other. The decoder then discovers all of the messages and the channel response when the number of samples taken for each received message are roughly greater than $(K+Slo{g}^{2}S)lo{g}^4(LN)$, and the number of messages is roughly at least $lo{g}^2(LN)$.

Cited in

(8)

• Blind three dimensional deconvolution via convex optimization
• Sensor calibration for off-the-grid spectral estimation
• Learning Iteration-wise Generalized Shrinkage-Thresholding Operators for Blind Deconvolution
• Self-calibration and bilinear inverse problems via linear least squares
• Manifold Gradient Descent Solves Multi-Channel Sparse Blind Deconvolution Provably and Efficiently
• Exact Recovery of Multichannel Sparse Blind Deconvolution via Gradient Descent
• Bisparse blind deconvolution through hierarchical sparse recovery
• BranchHull: convex bilinear inversion from the entrywise product of signals with known signs

This page was built for publication: Leveraging Diversity and Sparsity in Blind Deconvolution

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5375559)