Conference proceeding
Efficient Compressive Sensing with Deterministic Guarantees Using Expander Graphs
2007 IEEE Information Theory Workshop, pp.414-419
09/2007
DOI: 10.1109/ITW.2007.4313110
Abstract
Compressive sensing is an emerging technology which can recover a sparse signal vector of dimension n via a much smaller number of measurements than n. However, the existing compressive sensing methods may still suffer from relatively high recovery complexity, such as O (n 3 ), or can only work efficiently when the signal is super sparse, sometimes without deterministic performance guarantees. In this paper, we propose a compressive sensing scheme with deterministic performance guarantees using expander-graphs-based measurement matrices and show that the signal recovery can be achieved with complexity O (n) even if the number of nonzero elements k grows linearly with n. We also investigate compressive sensing for approximately sparse signals using this new method. Moreover, explicit constructions of the considered expander graphs exist. Simulation results are given to show the performance and complexity of the new method.
Details
- Title: Subtitle
- Efficient Compressive Sensing with Deterministic Guarantees Using Expander Graphs
- Creators
- Weiyu Xu - California Institute of TechnologyBabak Hassibi - California Institute of Technology
- Resource Type
- Conference proceeding
- Publication Details
- 2007 IEEE Information Theory Workshop, pp.414-419
- DOI
- 10.1109/ITW.2007.4313110
- Publisher
- IEEE
- Language
- English
- Date published
- 09/2007
- Academic Unit
- Electrical and Computer Engineering
- Record Identifier
- 9984197229202771
Metrics
19 Record Views