Frequency domain identification of complex sinusoids in the presence of additive noise

This paper describes a new approach for identifying the parameters of complex sinusoids from a finite number of measurements, in presence of additive and uncorrelated white noise. The proposed approach deals with frequency domain data and as a major feature, it enables the estimation to be frequency selective. In many aspects the new method resembles the well-known ESPRIT subspace algorithm, originally developed in the time domain. However, the sub-band frequency selective feature allows a reduction of the computations and can improve the quality of the estimates. The properties of the proposed method are analyzed by means of Monte Carlo simulations and its performance is compared with those of other estimation algorithms.