Impact of Statistically Small Population Sampling on Threshold Detection

This paper focuses on statistical sampling technique for the SNR-based threshold detection by using a small population sampling via cell averaging technique applied on a sliding wavelength window. We analyze and also demonstrate the threshold detection capability even with extremely small population sampling. This sampling technique is intended for fiber-optic transmission systems that share fiber Bragg gratings (FBG) sensing applications and telecommunication services in the same single mode fiber (SMF) simultaneously. This is demonstrated on two reference sets of data obtained by using two optical spectrum analyzers (OSA) with sparse and dense population samplings, respectively over the C-band fiber region. It has been shown that an appropriate sizing of statistical sampling has a potential to reduce demands on computational resources.