Constrained total least squares applied to frequency estimation of sinusoids
T.T. Abatzoglou, Victor C. Soon · 2003
Constrained total least squares (CTLS) is a technique for solving an overdetermined system of equations Ax=b that tries to remove the noise present in A and b while simultaneously exploiting the algebraic structure of the noise in A and b. The CTLS method is applied to the problem of estimating the frequencies of sinusoids in white noise. For the case of 25 samples, a sampling rate of 1 Hz, and two sinusoids whose frequency separation is 0.02, the resulting root-mean-square error of the CTLS technique is very close to that of the maximum likelihood estimator up to 3 dB signal-to-noise ratio.>