A novel adaptive filtering for LFM signal in FRFT domain
Authors:
- 1 October 2010
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 21645221,p. 239-242
Abstract
A novel adaptive filtering scheme based on Fractional Fourier Transform (FRFT) for time-varying linear frequency modulated (LFM) signal is introduced and characterized using the characteristic of good energy concentration on LFM signal in FRFT domain. Grading iterative calculation method is used to obtain the best rotation angle with low computational complexity. Leakage factor is introduced for adaptation to remove the memory effects and the step size is normalized by the power of input signals. In case of LFM signal in low SNR environment, the proposed normalized leakage LMS (NL-LMS) algorithm provides less error and faster convergence. In the end, effects of different parameters on convergence are studied and simulation results present the proposed algorithm has good convergence properties and small mean square error in fast time-varying and low SNR (SNR=-10dB) environment.Keywords
This publication has 7 references indexed in Scilit:
- Adaptive fractional Fourier domain filteringSignal Processing, 2010
- A Frequency Domain Step-Size Control Method for LMS AlgorithmsIEEE Signal Processing Letters, 2009
- Random Discrete Fractional Fourier TransformIEEE Signal Processing Letters, 2009
- Scheduled-Stepsize NLMS AlgorithmIEEE Signal Processing Letters, 2009
- Time-Delay Estimation of Chirp Signals in the Fractional Fourier DomainIEEE Transactions on Signal Processing, 2009
- LMS Coefficient Filtering for Time-Varying Chirped SignalsIEEE Transactions on Signal Processing, 2004
- Adaptive recovery of a chirped sinusoid in noise. II. Performance of the LMS algorithmIEEE Transactions on Signal Processing, 1991