Statistical model of a signal of Raman spectroscopy: Detection

Raman spectroscopy (RS) is a technique to find the spectral fingerprint of a material under study. To reach the spectrum, it is necessary to pass the acquired signal at the spectrometer by a bank of filters known as Raman signal's preprocessing system, which should eliminate all noise components accompanying the signal. The behavior of these noise components, and even the signal itself, are information that can be useful in order to optimize the system and to study the signal itself. This paper proposes a statistical model of Raman signal as the pre-processing system receives it, the signal's noise components and their statistical behavior, including its mathematical representation are presented. Additionally, it shows the results of the implementation of the simulation model leaving the door open for the validation, on which work is being done. The implementation of this model may be useful as an input signal for the optimization of the filter banks, since there are not always sufficient Raman signals for detailed study and development of filters for a specific signal under study. The model could also be used as a base for the study for systems using Raman spectroscopy to recognize substances.