Title :
Unbiased FIR system identification in the presence of input and output interference
Author_Institution :
Dept. of Electron. Eng., City Univ. of Hong Kong, Kowloon, China
Abstract :
In the presence of input interference, the Wiener solution for impulse response estimation is biased. In this paper, it is proved that bias removal can be achieved by proper scaling of the optimal filter coefficients and a modified least mean squares algorithm is then developed for accurate system identification in noise. Simulation results are included to compare the impulse response estimation performances of the proposed method and two total least squares based adaptive algorithms under different interference conditions
Keywords :
FIR filters; filtering theory; identification; interference (signal); least mean squares methods; noise; Wiener solution; filter coefficient scaling; impulse response estimation performance; input interference; interference conditions; modified LMS algorithm; modified least mean squares algorithm; noise; optimal filter coefficients; output interference; unbiased FIR system identification; Adaptive algorithm; Additive noise; Finite impulse response filter; Interference; Least mean square algorithms; Least squares approximation; Least squares methods; Noise measurement; System identification; Wiener filter;
Conference_Titel :
Circuits and Systems, 1999. 42nd Midwest Symposium on
Conference_Location :
Las Cruces, NM
Print_ISBN :
0-7803-5491-5
DOI :
10.1109/MWSCAS.1999.867733
