ALGORITHMS OF ACOUSTIC SIGNALS DEMODULATION IN FIBER-OPTIC COHERENT REFLECTOMETRIC SYSTEMS

Kuzmin Nikolay Nikitovich, Kachalov Sergey Konstantinocich, Zavjalov Anton Vladimirovich

This article describes the use of coherence reflectometers with distributed acoustic sensors (DAS) for the detection and analysis of vibroacoustic events, such as pipeline leaks, the movement of people and vehicles, digging and drilling of pipelines. Work is underway to expand the areas of their use, including the protection of maritime infrastructure. Technologies for the production of coherent reflectometers and RAD are well developed and are at the stage of mass production; work continues to improve their consumer characteristics, such as expanding the monitoring area, increasing accuracy and resolution, as well as improving the probabilistic characteristics of event detection and classification. The consumer characteristics of sensors are determined both by technical parameters (for example, the intrinsic noise of a coherent reflectometer, the sensitivity of the RAD) and by software and mathematics. The article discusses the use of modern adaptive algorithms for event detection under conditions of uncertainty, which will increase the monitoring area and improve accuracy and resolution. The results of combining classical logical classification algorithms with neural network algorithms are described, which makes it possible to improve the probabilistic characteristics of event classification.

Keywords:optical time domain reflectometry, signal demodulation equipment, fiber optic security equipment, coherent reflectometer, signal-to-noise ratio signal processing