Instantaneous frequency estimation using weighted least squares Tikhonov regularization with box constraints

Ahmed Sigiuk, Matthew J. Yedlin

In this report, we begin with a brief introduction into the concept of instantaneous frequency, where we use the complex analytical signal to obtain the time dependent signal phase, and differentiate the phase with respect to time to obtain the instantaneous frequency. We reformulate our problem into a weighted least squares Tikhonov regularization with box constraints, and explain the advantages of the added terms in helping to smooth and stabilize our results. We conclude with some synthetic and real data simulation results and compare the proposed algorithm results with a well established time-frequency distribution method.