© The Institution of Engineering and Technology
The authors propose a new approach for the adaptive identification of sparse systems. This approach improves on the recursive least squares (RLS) algorithm by adding a sparsity inducing weighted 1 norm penalty to the RLS cost function. Subgradient analysis is utilised to develop the recursive update equations for the calculation of the optimum system estimate, which minimises the regularised cost function. Two new algorithms are introduced by considering two different weighting scenarios for the 1 norm penalty. These new 1 relaxation-based RLS algorithms emphasise sparsity during the adaptive filtering process, and they allow for faster convergence than standard RLS when the system under consideration is sparse. The authors test the performance of the novel algorithms and compare it with standard RLS and other adaptive algorithms for sparse system identification.
References
-
-
1)
-
H. Chen ,
D. D ,
S. M
.
Atomic decomposition by basis pursuit.
SIAM J. Sci. Comput.
,
33 -
36
-
2)
-
Chen, Y., Gu, Y., Hero, A.O.: `Sparse LMS for system identification', Proc. ICASSP, 19–24 April 2009, p. 3125–3128.
-
3)
-
Asif, M.S., Romberg, J.: `Streaming measurements in compressive sensing: ', Proc. 42nd Asilomar Conf. on Signals, Systems and Computers, 26–29 October 2008, p. 1051–1058.
-
4)
-
Angelosante, D., Giannakis, G.B.: `RLS-weighted Lasso for adaptive estimation of sparse signals', Proc. ICASSP, 19–24 April 2009, p. 3245–3248.
-
5)
-
D. Angelosante ,
J. Bazerque ,
G. Giannakis
.
Online adaptive estimation of sparse signals: where RLS meets the 1-norm.
IEEE Trans. Signal Process.
,
7 ,
3436 -
3447
-
6)
-
M.A.T. Figueiredo ,
R.D. Nowak
.
An EM algorithm for wavelet-based image restoration.
IEEE Trans. Image Process.
,
906 -
916
-
7)
-
J. Jin ,
S. Mei
.
l0 norm constraint LMS algorithm for sparse system identification.
IEEE Signal Process. Lett.
,
9 ,
774 -
777
-
8)
-
Slavakis, K., Kopsinis, Y., Theodoridis, S.: `Adaptive algorithm for sparse system identification using projections onto weighted ', IEEE Int. Conf. on Acoustics Speech and Signal Processing (ICASSP), 2010, p. 3742–3745.
-
9)
-
M.H. Hayes
.
(1996)
Statistical digital signal processing and modeling.
-
10)
-
K. Skretting ,
K. Engan
.
Recursive least squares dictionary learning algorithm.
IEEE Trans. Signal Process.
,
4 ,
2121 -
2130
-
11)
-
Rao, B.D., Song, B.: `Adaptive filtering algorithms for promoting sparsity', Proc. ICASSP, 6–10 April 2003, 6, p. VI361–VI364.
-
12)
-
E.J. Candès ,
M.B. Wakin ,
S.P. Boyd
.
Enhancing sparsity by reweighted l1 minimization.
J. Fourier Anal. Appl.
,
877 -
905
-
13)
-
Eksioglu, E.M.: `RLS adaptive filtering with sparsity regularization', Tenth Int. Conf. on Information Sciences Signal Processing and their Applications (ISSPA), May 2010, p. 550–553.
-
14)
-
D.P. Bertsekas ,
A. Nedic ,
A.E. Ozdaglar
.
(2003)
Convex analysis and optimization.
-
15)
-
R. Tibshirani
.
Regression shrinkage and selection via the lasso.
J. R. Stat. Soc. B
,
1 ,
267 -
288
-
16)
-
D.L. Duttweiler
.
Proportionate normalized least-mean-squares adaptation in echo cancellers.
IEEE Trans. Speech Audio Process.
,
5 ,
508 -
518
-
17)
-
B. Babadi ,
N. Kalouptsidis ,
V. Tarokh
.
SPARLS: the sparse RLS algorithm.
IEEE Trans. Signal Process.
,
8 ,
4013 -
4025
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-spr.2010.0083
Related content
content/journals/10.1049/iet-spr.2010.0083
pub_keyword,iet_inspecKeyword,pub_concept
6
6