Accession Number:



Deterministic Compressed Sensing

Descriptive Note:

Doctoral thesis

Corporate Author:


Personal Author(s):

Report Date:


Pagination or Media Count:



The central goal of compressed sensing is to capture attributes of a signal using very few measurements. The initial publications by Donoho and by Candes and Tao have been followed by applications to image compression, data streaming, medical signal processing, digital communication and many others. The emphasis has been on random sensing but the limitations of this framework include performance guarantees storage requirements, and computational cost. This thesis will describe two deterministic alternatives. The first alternative is based on expander graphs. We first show how expander graphs are appropriate for compressed sensing in terms of providing explicit and efficient sensing matrices as well as simple and efficient recovery algorithms. We show that by reformulating signal reconstruction as a zero-sum game we can efficiently recover any sparse vector. We provide a saddle-point reformulation of the expander-based sparse approximation problem, and propose an efficient expander-based sparse approximation algorithm, called the GAME algorithm. We show that the restricted isometry property of expander matrices in the l1-norm ensures that the GAME algorithm always recovers a sparse approximation to the optimal solution with an l1l1 data-domain approximation guarantee. We also demonstrate resilience to Poisson noise. The Poisson noise model is appropriate for a variety of applications, including low-light imaging and digital streaming where the signal-independent andor bounded noise models used in the compressed sensing literature are no longer applicable. We develop a novel sensing paradigm based on expander graphs and propose a MAP algorithm for recovering sparse or compressible signals from Poisson observations. We support our results with experimental demonstrations of reconstructing average packet arrival rates and instantaneous packet counts at a router in a communication network, where the arrivals of packets in each flow follow a Poisson process.

Subject Categories:

  • Cybernetics
  • Miscellaneous Detection and Detectors

Distribution Statement: