Study of Frequency domain exponential functional link network filters
Authors:
T. Yu,
S. Tana,
R. C. de Lamareb,
Y. Yu
Abstract:
The exponential functional link network (EFLN) filter has attracted tremendous interest due to its enhanced nonlinear modeling capability. However, the computational complexity will dramatically increase with the dimension growth of the EFLN-based filter. To improve the computational efficiency, we propose a novel frequency domain exponential functional link network (FDEFLN) filter in this paper.…
▽ More
The exponential functional link network (EFLN) filter has attracted tremendous interest due to its enhanced nonlinear modeling capability. However, the computational complexity will dramatically increase with the dimension growth of the EFLN-based filter. To improve the computational efficiency, we propose a novel frequency domain exponential functional link network (FDEFLN) filter in this paper. The idea is to organize the samples in blocks of expanded input data, transform them from time domain to frequency domain, and thus execute the filtering and adaptation procedures in frequency domain with the overlap-save method. A FDEFLN-based nonlinear active noise control (NANC) system has also been developed to form the frequency domain exponential filtered-s least mean-square (FDEFsLMS) algorithm. Moreover, the stability, steady-state performance and computational complexity of algorithms are analyzed. Finally, several numerical experiments corroborate the proposed FDEFLN-based algorithms in nonlinear system identification, acoustic echo cancellation and NANC implementations, which demonstrate much better computational efficiency.
△ Less
Submitted 11 January, 2022;
originally announced January 2022.
