Parameter estimation for exponential data modeling of electro-magnetic fields

Well established estimation algorithms for exponential data models, such as GPOF or MV-HTLS, provide good results under two conditions: The available signals are long compared to their periods and the parameters are constant within the window. This paper demonstrates how signal parameters can be estimated when both conditions are seriously violated. Three different algorithms are outlined and compared to other, frequently used methods and the corresponding Cramer-Rao bound. Simulation results are presented.