Parameter estimation of AM and FM signals in noise

The problem of estimating the amplitude and frequencies of sinusoids in noise is a classical problem in signal processing. Subspace techniques which have a direct connection with the Hankel based realization problem, are known to provide asymptotically unbiased estimates of the signal parameters. These subspace techniques are also robust to noise in the data due to the signal and noise decomposition property of the singular value decomposition. We present a more general problem, i.e., that of estimating the amplitude and frequencies of amplitude modulation (AM) and frequency modulation (FM) signals in noise. We show that the solution is similar to that of the sinusoids in noise problem in both cases and show how to determine the parameters