Weak signal measurement via an accurate cross-spectral SVD-TLS approach

This paper describes an accurate cross-spectral Singular Value Decomposition-Total Least Squares (SVD-TLS) method to measure the frequencies of weak sinusoidal signals in measurement noise. The new algorithm can overcome efficiently the shortcomings of low resolution caused by characteristic equation extention and distinguish perfectly the false peaks from the harmonic signal peaks in power spectrum figures. Simulation results show that this new cross-spectral SVD-TLS algorithm can eliminate spectral false peaks and have high resolution