Pattern recognition algorithm and software design of an optical fiber vibration signal based on Φ-optical time-domain reflectometry

A phase-sensitive optical time-domain reflectometry with a detection distance of 10 km is used in this paper to recognize and to detect in real time events along the perimeter security monitoring system. To optimize the signal processing in the software system and achieve the design of distributed optical fiber vibration signal of the perimeter security monitoring system, a two-level vibration pattern recognition scheme based on the power spectral estimation is proposed. The monitoring system software of the proposed algorithm is designed and implemented, as well as the functional modules involved in the design of the software system. The experimental tests of the whole system show that the proposed algorithm can effectively identify in real time any intrusion event along the sensing cable, with an intrusion detection rate greater than 95%, including false alarms. Also, the corresponding monitoring system software can display the data collected in real time, with a processing time of around 1.6 s, which proves its reliability, thereby ensuring the system's validity. (C) 2019 Optical Society of America