An algorithm for efficient estimation of superimposed exponential signals
Zhidong Bai, C. R. Rao, Mosuk Chow · 2003
A computational algorithm is given for obtaining asymptotically efficient estimates of the unknown complex amplitudes and frequencies in a superimposed exponential model for signals. It is shown that the variance covariance matrix of these estimates is asymptotically the same as that for the maximum likelihood estimates, and thus the lower bound can be obtained.>