DocumentCode :
3065527
Title :
The dynamics of message passing on dense graphs, with applications to compressed sensing
Author :
Bayati, Mohsen ; Montanari, Andrea
Author_Institution :
Dept. of Electr. Eng., Stanford Univ., Stanford, CA, USA
fYear :
2010
fDate :
13-18 June 2010
Firstpage :
1528
Lastpage :
1532
Abstract :
`Approximate message passing´ algorithms proved to be extremely effective in reconstructing sparse signals from a small number of incoherent linear measurements. Extensive numerical experiments further showed that their dynamics is accurately tracked by a simple one-dimensional iteration termed state evolution. In this paper we provide the first rigorous foundation to state evolution. We prove that indeed it holds asymptotically in the large system limit for sensing matrices with iid gaussian entries. While our focus is on message passing algorithms for compressed sensing, the analysis extends beyond this setting, to a general class of algorithms on dense graphs. In this context, state evolution plays the role that density evolution has for sparse graphs.
Keywords :
Gaussian processes; graph theory; iterative methods; matrix algebra; message passing; signal reconstruction; compressed sensing; dense graphs; iid Gaussian entry; incoherent linear measurements; message passing algorithm; one-dimensional iteration termed state evolution; sensing matrices; sparse graphs; sparse signal reconstruction; Algorithm design and analysis; Approximation algorithms; Compressed sensing; Electric variables measurement; Message passing; Numerical simulation; Sparse matrices; Statistics; Tree graphs; Vectors;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information Theory Proceedings (ISIT), 2010 IEEE International Symposium on
Conference_Location :
Austin, TX
Print_ISBN :
978-1-4244-7890-3
Electronic_ISBN :
978-1-4244-7891-0
Type :
conf
DOI :
10.1109/ISIT.2010.5513529
Filename :
5513529
Link To Document :
بازگشت