© The Institution of Engineering and Technology
In this Letter, the authors present an improved sparse recursive least squares (RLS) algorithm, which employs a novel approximation of the norm of the filter coefficient vector for regularising the RLS cost function. The proposed algorithm achieves improved performance over existing algorithms as demonstrated via numerical simulations.
References
-
-
1)
-
1. Eksioglu, E.M., Tanc, A.K.: ‘RLS algorithm with convex regularization’, Signal Process. Lett., 2011, 18, (8), pp. 470–473 (doi: 10.1109/LSP.2011.2159373).
-
2)
-
46. Angelosante, D., Bazerque, J.A., Giannakis, G.B.: ‘Online adaptive estimation of sparse signals: where RLS meets the ll1-norm’, IEEE Trans. Signal Process., 2010, 58, (7), pp. 3436–3447 (doi: 10.1109/TSP.2010.2046897).
-
3)
-
6. Babadi, B., Kalouptsidis, N., Tarokh, V.: ‘SPARLS: the sparse RLS algorithm’, Trans. Signal Proc., 2010, 58, (8), pp. 4013–4025 (doi: 10.1109/TSP.2010.2048103).
-
4)
-
4. Horn, R., Johnson, C.R.: ‘Matrix analysis’ (Cambridge University, Cambridge, 1985).
-
5)
-
3. Hong, X., Gao, J., Chen, S.: ‘Zero attracting recursive least squares algorithms’, Trans. Veh. Technol., 2017, 66, (1), pp. 213–221.
-
6)
-
24. Su, G., Jin, J., Gu, Y., et al: ‘Performance analysis of l0-norm constraint least mean square algorithm’, IEEE Trans. Signal Proecess., 2012, 60, (5), pp. 2223–2235 (doi: 10.1109/TSP.2012.2184537).
http://iet.metastore.ingenta.com/content/journals/10.1049/el.2017.3441
Related content
content/journals/10.1049/el.2017.3441
pub_keyword,iet_inspecKeyword,pub_concept
6
6