Abstract
Compressive Sensing (CS) is an emerging research area that allows efficient signal acquisition under the sub-Nyquist rate while still promising reliable data recovery. However, practical applications of CS in hardware platforms are limited as signal reconstruction is still challenging due to its high computational complexity, especially for autonomous real-time signal recovery. In this article, we propose an algorithmic transformation technique referred to as Matrix Inversion Bypass (MIB) to improve the signal recovery efficiency of the Orthogonal Matching Pursuit (OMP)-based CS reconstruction. The basic idea of MIB is to decouple the computations of intermediate signal estimates and matrix inversions, thereby enabling parallel processing of these two time-consuming operations in the OMP algorithm. The proposed MIB naturally leads to a parallel architecture for high-speed dedicated hardware implementations. An FPGA-based implementation is developed with the optimized structure aimed at the efficient utilization of hardware resources while realizing high-speed signal recovery. The proposed architecture can perform the signal recovery at up to 1.4 × faster than the OMP-based implementation using almost the same hardware resources.
- Nasir Ahmed, T. Natarajan, and Kamisetty Rao. 1974. Discrete cosine transform. IEEE Trans. Comput. C-23, 1 (Jan. 1974), 90--93. Google Scholar
Digital Library
- Lin Bai, Patrick Maechler, Michael Muehlberghuber, and Hubert Kaeslin. 2012. High-speed compressed sensing reconstruction on FPGA using OMP and AMP. In Proceedings of 2012 19th IEEE International Conference on Electronics, Circuits and Systems, 53--56.Google Scholar
Cross Ref
- Waheed Bajwa, Jarvis Haupt, Akbar Sayeed, and Robert Nowak. 2012. Compressed channel sensing: A new approach to estimating sparse multipath channels. Proc. IEEE 98, 6 (June 2010), 1058--1076.Google Scholar
- Dror Baron, Marco Duarte, Michael Wakin, Shriram Sarvotham, and Richard Baraniuk. 2009. Distributed compressive sensing. arXiv preprint arXiv:0901.3403 (Jan. 2009).Google Scholar
- Christian Berger, Zhaohui Wang, Jianzhong Huang, and Shengli Zhou. 2010. Application of compressive sensing to sparse channel estimation. IEEE Commun. Mag. 48, 11 (Nov. 2010), 164--174. Google Scholar
Digital Library
- Ake Björck. 1996. Numerical Methods for Least Squares Problems. SIAM.Google Scholar
- Emmanuel J. Candès. 2006. Compressive sampling. In Proc. Int. Congr. Math. 3, 12 (Aug. 2006), 1433--1452.Google Scholar
- Emmanuel Candès and Terence Tao. 2005. Decoding by linear programming. IEEE Trans. Inf. Theory 51, 12 (Dec. 2005), 4203--4215. Google Scholar
Digital Library
- Emmanuel Candès and Michael Wakin. 2008. An introduction to compressive sampling. IEEE Signal Process. Mag. 25, 2 (Mar. 2008), 21--30.Google Scholar
Cross Ref
- Krishnendu Chakrabarty, Sitharama Iyengar, Hairong Qi, and Eungchun Cho. 2002. Grid coverage for surveillance and target location in distributed sensor networks. IEEE Trans. Comput 51, 12 (Dec. 2002), 1448--1453. Google Scholar
Digital Library
- Scott Chen, David Donoho, and Michael Saunders. 2001. Atomic decomposition by basis pursuit. SIAM Rev. 43, 1 (Mar. 2001), 129--159. Google Scholar
Digital Library
- Wei Dai and Olgica Milenkovic. 2009. Subspace pursuit for compressive sensing signal reconstruction. IEEE Trans. Inf. Theory 55, 5 (May 2009), 1094--1121. Google Scholar
Digital Library
- David Donoho. 2006. Compressed sensing. IEEE Trans. Inf. Theory 52, 4 (Apr. 2006), 1289--1306. Google Scholar
Digital Library
- David Donoho, Yaakov Tsaig, Iddo Drori, and Jean-Luc Starck. 2012. Sparse solution of underdetermined systems of linear equations by stagewise orthogonal matching pursuit. IEEE Trans. Inf. Theory 58, 2 (Feb. 2012), 1094--1121. Google Scholar
Digital Library
- Marco Duarte, Mark Davenport, Dharmpal Takhar, Jason Laska, Ting Sun, Kevin Kelly, and Richard Baraniuk. 2008. Single-pixel imaging via compressive sampling. IEEE Signal Process. Mag. 25, 2 (Mar. 2008), 83--91.Google Scholar
Cross Ref
- Li Feng, Robert Grimm, Kai Block, Hersh Chandarana, Sungheon Kim, Jian Xu, L. Axel, D. K. Sodickson, and R. Otazo. 2014. Golden-angle radial sparse parallel MRI: Combination of compressed sensing, parallel imaging, and golden-angle radial sampling for fast and flexible dynamic volumetric MRI. Magn. Reson. Med. 72, 3 (Sep. 2014), 707--717.Google Scholar
Cross Ref
- Mário Figueiredo, Robert Nowak, and Stephen Wright. 2007. Gradient projection for sparse reconstruction: Application to compressed sensing and other inverse problems. IEEE J. Sel. Topics Signal Process. 1, 4 (Dec. 2007), 586--597.Google Scholar
Cross Ref
- Gene Golub and Charles Van Loan. 1983. Matrix Computations. Johns Hopkins University Press, Baltimore.Google Scholar
- Justin Haldar, Diego Hernando, and Zhi-Pei Liang. 2011. Compressed-sensing MRI with random encoding. IEEE Trans. Med. Imag. 30, 4 (Apr. 2011), 893--903.Google Scholar
Cross Ref
- Sitharama Iyengar and Richard Brooks (Eds.). 2005. Distributed sensor networks. CRC Press.Google Scholar
- Stefan Kunis and Holger Rauhut. 2008. Random sampling of sparse trigonometric polynomials, II. Orthogonal matching pursuit versus basis pursuit. J. Found. Comput. Math. 8, 6 (Dec. 2008), 737--763. Google Scholar
Digital Library
- Yun-Yueh Lee, Ching-Hung Wang, and Yuan-Hao Huang. 2015. A hybrid RF/baseband precoding processor based on parallel-index-selection matrix-inversion-bypass simultaneous orthogonal matching pursuit for millimeter wave MIMO systems. IEEE Trans. Signal Process. 63, 2 (Jan. 2015), 305--317.Google Scholar
Cross Ref
- Hossein Mamaghanian, Nadia Khaled, David Atienza, and Pierre Vandergheynst. 2012. Design and exploration of low-power analog to information conversion based on compressed sensing. IEEE Trans. Emerg. Sel. Topics Circuits Syst. 2, 3 (Sep. 2012), 493--501.Google Scholar
Cross Ref
- Guoxian Huang and Lei Wang. 2012. High-speed signal reconstruction with orthogonal matching pursuit via matrix inversion bypass. In Proceedings of the 2012 IEEE Workshop on Signal Processing Systems (SiPS’12), 191--196. Google Scholar
Digital Library
- Oleg Maslennikow, Volodymyr Lepekha, Anatoli Sergiyenko, Adam Tomas, and Roman Wyrzykowski. 2008. Parallel implementation of Cholesky LLT-algorithm in FPGA-based processor. Parallel Process. Appl. Math. 4967 (2008), 137--147. Google Scholar
Digital Library
- Mark Murphy, Marcus Alley, James Demmel, Kurt Keutzer, Shreyas Vasanawala, and Michael Lustig. 2012. Fast ℓ1-SPIRiT compressed sensing parallel imaging MRI: Scalable parallel implementation and clinically feasible runtime. IEEE Trans. Med. Imag. 31, 6 (June 2012), 1250--1262.Google Scholar
Cross Ref
- Balas Natarajan. 1995. Sparse approximate solutions to linear systems. SIAM J. Comput. 24, 2 (Apr. 1995), 227--234. Google Scholar
Digital Library
- Deanna Needell and Roman Vershynin. 2009. Uniform uncertainty principle and signal recovery via regularized orthogonal matching pursuit. Found. Comput. Math. 9, 3 (Apr. 2009), 317--334. Google Scholar
Digital Library
- Kangyu Ni, Prasun Mahanti, Somantika Datta, Svetlana Roudenko, and Douglas Cochran. 2009. Image reconstruction by deterministic compressed sensing with chirp matrices. Proc. SPIE 7497 (Oct. 2009), 74971S--74971S.Google Scholar
Cross Ref
- Yusuke Oike and Abbas Gamal. 2013. CMOS image sensor with per-column ΣΔ ADC and programmable compressed sensing. IEEE J. Solid-State Circuits 48, 1 (Jan. 2013), 318--328.Google Scholar
Cross Ref
- Hassan Rabah, Abbes Amira, Basant Kumar Mohanty, Somaya Almaadeed, and Pramod Kumar Meher. 2015. FPGA implementation of orthogonal matching pursuit for compressive sensing reconstruction. in IEEE Trans. Very Large Scale Integr. (VLSI) Syst. (Oct. 2015), 2209--2220.Google Scholar
- Fengbo Ren, Richard Dorrace, Wenyao Xu, and Dejan Markovió. 2013. A single-precision compressive sensing signal reconstruction engine on FPGAs. In Proceedings of the 2013 23rd International Conference on Field Programmable Logic and Applications. 1--4.Google Scholar
Cross Ref
- Ryan Robucci, Jordan Gray, Leung Chiu, Justin Romberg, and Paul Hasler. 2010. Compressive sensing on a CMOS separable-transform image sensor. Proc. IEEE 98, 6 (June 2010), 1089--1101.Google Scholar
Cross Ref
- Ron Rubinstein, Michael Zibulevsky, and Michael Elad. 2008. Effiecient implementation of the K-SVD algorithm using batch orthogonal matching pursuit. CS Technion 40, 8 (Apr. 2008), 1--15.Google Scholar
- Avi Septimus and Raphael Steinberg. 2010. Compressive sampling hardware reconstruction. In Proceedings of 2010 IEEE International Symposium on Circuits and Systems. 3316--3319.Google Scholar
Cross Ref
- John Slavinsky, Jason Laska, Mark Davenport, and Richard Baraniuk. 2011. The compressive multiplexer for multi-channel compressive sensing. IEEE Int. Conf. Acoust., Speech, Signal Process. 31, 6 (May 2011), 3980--3983.Google Scholar
Cross Ref
- Jerome Stanislaus and Tinoosh Mohsenim. 2012. High performance compressive sensing reconstruction hardware with QRD process. In Proceedings of 2012 International Symposium on Circuits and Systems, 29--32.Google Scholar
Cross Ref
- Bob Sturm and Mads Christensen. 2015. Comparison of orthogonal matching pursuit implementations. In Proceedings of EUSIPCO. 220--224.Google Scholar
- Joel Tropp and Anna Gilbert. 2007. Signal recovery from random measurements via orthogonal matching pursuit. IEEE Trans. Inf. Theory 53, 12 (Dec. 2007), 4655--4666. Google Scholar
Digital Library
- Joel Tropp, Jason Laska, Marco Duarte, Justin Romberg, and Richard Baraniuk. 2010. Beyond Nyquist: Efficient sampling of sparse bandlimited signals. IEEE Trans. Inf. Theory 56, 1 (Jan. 2010), 520--544. Google Scholar
Digital Library
Index Terms
An FPGA-Based Architecture for High-Speed Compressed Signal Reconstruction
Recommendations
High-Speed Signal Reconstruction with Orthogonal Matching Pursuit via Matrix Inversion Bypass
SIPS '12: Proceedings of the 2012 IEEE Workshop on Signal Processing SystemsCompressive sensing (CS) is an emerging research area that has great significance to the design of resource-constrained cyber physical systems. Signal reconstruction in CS remains a challenge due to its high computational complexity, which limits the ...
New conditions for uniformly recovering sparse signals via orthogonal matching pursuit
Recently, lots of work has been done on conditions of guaranteeing sparse signal recovery using orthogonal matching pursuit (OMP). However, none of the existing conditions is both necessary and sufficient in terms of the so-called restricted isometric ...
Subspace pursuit for compressive sensing signal reconstruction
We propose a new method for reconstruction of sparse signals with and without noisy perturbations, termed the subspace pursuit algorithm. The algorithm has two important characteristics: low computational complexity, comparable to that of orthogonal ...






Comments