Vector Approximate Message Passing Algorithm for Structured Perturbed Sensing Matrix
arXiv:1808.08579
Abstract
In this paper, we consider a general form of noisy compressive sensing (CS) where the sensing matrix is not precisely known. Such cases exist when there are imperfections or unknown calibration parameters during the measurement process. Particularly, the sensing matrix may have some structure, which makes the perturbation follow a fixed pattern. While previous work has focused on extending the approximate message passing (AMP) and LASSO algorithm to deal with the independent and identically distributed (i.i.d.) perturbation, we propose the robust variant vector approximate message passing (VAMP) algorithm with the perturbation being structured, based on the recent VAMP algorithm. The performance of the robust version of VAMP is demonstrated numerically.
arXiv comment6 pages, 3 figures
References in corpus (10)
- Message Passing Algorithms for Compressed Sensing
- Expectation-Maximization Gaussian-Mixture Approximate Message Passing
- Sparsity-Cognizant Total Least-Squares for Perturbed Compressive Sampling
- General Deviants: An Analysis of Perturbations in Compressed Sensing
- Sparse recovery under matrix uncertainty
- A Message-Passing Receiver for BICM-OFDM over Unknown Clustered-Sparse Channels
- Perturbation Analysis of Orthogonal Matching Pursuit
- Hybrid Approximate Message Passing
- Generalized Approximate Message Passing for Estimation with Random Linear Mixing
- Compressed Sensing under Matrix Uncertainty: Optimum Thresholds and Robust Approximate Message Passing