FIR filter for frequency estimation using eigen-approach
F.B. Elissa, Mohammad Jamil Mismar · 2017 International Conference on Electrical and Computing Technologies and Applications (ICECTA) · 2017
A new method for estimating the frequency of sinusoidal signals is proposed using Eigen analysis. The eigenvector of the correlation matrix that sets the FIR filter's output power to a minimum value corresponds to the minimum eigen value. Moreover, the filter polynomial, with coefficients as the eigenvector, determines the corresponding roots. The pseudospectrum is determined by root elimination and output power estimation. The frequency, pseudo-spectrum, and number of source signals are all estimated.