A unified minimum variance spectrum-based approach for simultaneousidentification of both harmonic and stationary random noise fields

The technique presented in this correspondence uses the MV spectrum's convergence properties to identify unknown and arbitrary harmonic signal fields. The correlation sequence for the identified harmonic signal field is then combined with the observed overall correlation sequence to obtain a spectral model of the random noise field