TY - JOUR
T1 - VLSI design of approximate message passing for signal restoration and compressive sensing
AU - Maechler, Patrick
AU - Studer, Christoph
AU - Bellasi, David E.
AU - Maleki, Arian
AU - Burg, Andreas
AU - Felber, Norbert
AU - Kaeslin, Hubert
AU - Baraniuk, Richard G.
N1 - Funding Information:
Manuscript received February 29, 2012; revised July 03, 2012; accepted August 02, 2012. Date of publication October 16, 2012; date of current version December 05, 2012. The work of C. Studer was supported by the Swiss National Science Foundation (SNSF) under Grant PA00P2-134155. The work of A. Maleki and R. G. Baraniuk was supported in part by the National Science Foundation (NSF) under Grant CCF-0431150, Grant CCF-0728867, and Grant CCF-0926127, in part by the Defense Advanced Research Projects Agency Office of Naval Research (DARPA/ONR) under Grant N66001-08-1-2065, Grant N66001-11-1-4090, Grant N66001-11-C-4092, in part by the ONR under Grant N00014-08-1-1112 and Grant N00014-10-1-0989, in part by the Air Force Office of Scientific Research (AFOSR) under Grant FA9550-09-1-0432, Army Research Office Multidisciplinary University Research Initiative (ARO MURI) under Grant W911NF-07-1-0185 and Grant W911NF-09-1-0383, and in part by the Texas Instruments Leadership University Program. This paper was recommended by Guest Editor R. Rovatti.
Copyright:
Copyright 2013 Elsevier B.V., All rights reserved.
PY - 2012
Y1 - 2012
N2 - Sparse signal recovery finds use in a variety of practical applications, such as signal and image restoration and the recovery of signals acquired by compressive sensing. In this paper, we present two generic very-large-scale integration (VLSI) architectures that implement the approximate message passing (AMP) algorithm for sparse signal recovery. The first architecture, referred to as AMP-M, employs parallel multiply-accumulate units and is suitable for recovery problems based on unstructured (e.g., random) matrices. The second architecture, referred to as AMP-T, takes advantage of fast linear transforms, which arise in many real-world applications. To demonstrate the effectiveness of both architectures, we present corresponding VLSI and field-programmable gate array implementation results for an audio restoration application. We show that AMP-T is superior to AMP-M with respect to silicon area, throughput, and power consumption, whereas AMP-M offers more flexibility.
AB - Sparse signal recovery finds use in a variety of practical applications, such as signal and image restoration and the recovery of signals acquired by compressive sensing. In this paper, we present two generic very-large-scale integration (VLSI) architectures that implement the approximate message passing (AMP) algorithm for sparse signal recovery. The first architecture, referred to as AMP-M, employs parallel multiply-accumulate units and is suitable for recovery problems based on unstructured (e.g., random) matrices. The second architecture, referred to as AMP-T, takes advantage of fast linear transforms, which arise in many real-world applications. To demonstrate the effectiveness of both architectures, we present corresponding VLSI and field-programmable gate array implementation results for an audio restoration application. We show that AMP-T is superior to AMP-M with respect to silicon area, throughput, and power consumption, whereas AMP-M offers more flexibility.
KW - Approximate message passing (AMP)
KW - compressive sensing (CS)
KW - ell-norm minimization
KW - fast discrete cosine transform (DCT)
KW - field-programmable gate array (FPGA)
KW - signal restoration
KW - sparse signal recovery
KW - very-large scale integration (VLSI)
UR - http://www.scopus.com/inward/record.url?scp=84870983489&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84870983489&partnerID=8YFLogxK
U2 - 10.1109/JETCAS.2012.2214636
DO - 10.1109/JETCAS.2012.2214636
M3 - Article
AN - SCOPUS:84870983489
VL - 2
SP - 579
EP - 590
JO - IEEE Journal on Emerging and Selected Topics in Circuits and Systems
JF - IEEE Journal on Emerging and Selected Topics in Circuits and Systems
SN - 2156-3357
IS - 3
M1 - 6331565
ER -