Abstract
The present paper reviews a common problem to many different scientific disciplines: the fitting of multiexponential time responses. Several algorithms have been carried out in order to solve this kind of problems, but all of them deal with the inherent difficulty in overcoming noise sources. In effect, the ill-conditioning of the problem restricts the accuracy of the results. Accordingly, we present a comparative study of time domain algorithms for fitting multiexponential responses. Four algorithms that can be found in the literature have been chosen for this purpose. The work has been particularly focused on the case of thermally activated processes that we have considered as an adequate touchstone. The analysis has been carried out by computer simulation. The same working conditions of signal-to-noise ratio and response complexity has been applied to each algorithm. The results are compared and the origin of discrepancies between these results and the exact solutions are discussed.