Performance investigation of real-time fiber optic perimeter intrusion detection systems using event classification

Abstract

The success of any perimeter intrusion detection system depends on three important performance parameters: the probability of detection (POD), the nuisance alarm rate (NAR), and the false alarm rate (FAR). The most fundamental parameter, POD, is normally related to a number of factors such as the event of interest, the sensitivity of the sensor, the installation quality of the system and the reliability of the sensing equipment. A critical performance parameter of any outdoor perimeter intrusion detection system is its capability of discriminating between intrusion and nuisance events without compromising its sensitivity or POD. In this paper, the performance criteria of a real-time fence-mounted distributed fiber-optic intrusion detection system are discussed. The use of event recognition and classification techniques to maintain high detection rates and minimize alarms caused by nuisance events is presented with a performance analysis for different event classification algorithms. Practical results for the detection of intrusion events and suppression of nuisances are also shown.