Amplitude estimation of sinusoidal signals: survey, new results, and an application

摘要: This paper considers the problem of amplitude estimation sinusoidal signals from observations corrupted by colored noise. A relatively large number estimators, which encompass least squares (LS) and weighted (WLS) methods, are described. Additionally, filterbank approaches, widely used for spectral analysis, extended to estimation; more exactly, we consider matched-filterbank (MAFI) approach show that appropriately designing prefilters, MAFI includes WLS approach. The techniques discussed in this do not model observation noise, yet, they all asymptotically statistically efficient. It is, however, their different finite-sample properties particular interest study. Numerical examples provided illustrate differences among various estimators. Although applications numerous, focus herein on system identification using probing provide a computationally simple accurate solution.