Self-Contained Reference Sensors to Reduce Nuisance Alarm Rate in φ-OTDR-Based Fence Intrusion Detection System

Nuisance alarm rate (NAR) is one of the key performance parameters in a phase-sensitive optical time domain reflectometry (φ-OTDR)-based fence intrusion detection system. Typically, the vibrations caused by ambient environmental conditions, such as heavy rain, strong wind, and passing vehicles, easily result in many nuisance alarms. Significant research efforts have been undertaken to suppress the NAR. In this paper, we propose to utilize short segments of the sensing fiber as reference sensors for significant reduction in the NAR in φ-OTDR for the first time, to the best of our knowledge. According to our field trial results, the proposed approach can reduce the NAR by more than 90%. The proposed approach is very simple, practical, and cost-effective, which can be easily integrated with the existing methods of reducing NAR and act as an additional level of decision-making algorithm for triggering alarms.