Abstract
The least-squares curve fitting procedure of experimental data is described for heterodyne detection. Two independent methods, the Maximum Likelihood method of Fisher and SIMPLEX, have been tried. The relative merits and limitations are discussed in detail.