Vibration Pattern Recognition and Classification in OTDR Based Distributed Optical-Fiber Vibration Sensing System

In this paper we propose and demonstrate the scheme of vibration pattern recognition and classification in the OTDR based distributed optical-fiber vibration sensing system. We set up the engineering system with signal processing PC for perimeter security in some high-tech park in Nanjing. Three types of disturbing actions, including climbing up and kicking at the wall by a person, and watering on the sensing optical fiber cable same as the rain falling on, are implemented. By using level crossing rate (LCR), we can obtain their individual pattern features, so that the eigenvalue database for three disturbing actions can be built in the system. By comparing three types of vibrations, the differences among these can be given out. The results show three vibration patterns can be recognized and classified effectively.