Transient multiexponential signals analysis using Bayesian deconvolution

A new method based on Bayesian deconvolution is proposed for multiexponential transient signal analysis. The multiexponential signal is initially converted to a convolution model using logarithmic and differential transformation after which the Bayesian iteration is used to deconvolve the data. The numerical simulation is applied on four different multiexponential signals with different levels of noise. Thermal transient experiment data of the high power light emitting diodes are also analyzed using the proposed method. Simulation and experimental results indicate that the present method performs efficiently in accurately estimating the decay rates except at low SNR case.