Title :
Exponential normalized sign algorithm for system identification
Author :
Jeong, Jae Jin ; Kim, Sang Woo
Author_Institution :
Div. of Electr. & Comput. Eng., Pohang Univ. of Sci. & Technol. (POSTECH), Pohang, South Korea
Abstract :
The sign algorithm is not influenced by impulsive noise, but it exhibits slow convergence rate. Therefore, in this paper, the exponential normalized sign algorithm is proposed for system identification. The proposed algorithm is obtained from the exponential function of the L1 norm of the output error signal. The results of computer simulations proved that the convergence rate of the proposed algorithm is better than that of the sign algorithm.
Keywords :
adaptive filters; convergence; identification; impulse noise; L1 norm; adaptive filters; computer simulations; convergence rate; exponential function; exponential normalized sign algorithm; impulsive noise; output error signal; system identification; Algorithm design and analysis; Convergence; Noise; Signal processing algorithms; Simulation; Vectors; Adaptive filtering; Colored input singal; Impulsive noise; Outlier; Stochastic gradient method;
Conference_Titel :
Information, Communications and Signal Processing (ICICS) 2011 8th International Conference on
Conference_Location :
Singapore
Print_ISBN :
978-1-4577-0029-3
DOI :
10.1109/ICICS.2011.6174219
