Impact of Reducing Statistically Small Population Sampling on Threshold Detection in FBG Optical Sensing

Area of Science:

• Optical Fiber Communications
• Signal Processing
• Statistical Analysis

Background:

• Optical fiber systems transmit diverse data, necessitating efficient information recovery techniques.
• Fiber Bragg Gratings (FBGs) are crucial for sensing and communication, requiring accurate spectral analysis.
• Statistical population sampling offers potential for optimizing data retrieval from complex signals.

Purpose of the Study:

• To develop a simple statistical sampling method for efficient information retrieval from optical fiber.
• To investigate statistical denoising and detection of Fiber Bragg Grating (FBG) power spectra.
• To analyze the impact of sliding window techniques and small population sampling on data recovery.

Main Methods:

• Investigated one-sided and two-sided sliding window techniques for FBG power spectra analysis.
• Employed small population sampling for statistical denoising and threshold detection.
• Experimentally varied window sizes up to half the FBG power spectra bandwidth.
• Analyzed detection thresholds based on mean and standard deviation of sampled data.

Main Results:

• Shorter sliding windows significantly reduce processing latency, benefiting real-time applications.
• The normality three-sigma rule is not essential for small population sampling in FBG analysis.
• Novel denoising and detection methods are independent of prior probability distribution knowledge.
• Detection threshold adaptability is strongly correlated with the mean and standard deviation of the sampled population.

Conclusions:

• A simplified statistical sampling approach enables efficient FBG spectral analysis in optical fibers.
• The method offers reduced latency and robust performance without requiring prior data distribution information.
• This technique is adaptable and suitable for real-time signal processing in optical communication and sensing systems.