Fence vibrations [intruder detection

Abstract

Many Perimeter Intrusion Detection Systems (PIDS) use microphonic sensors mounted on fences to detect intruder attacks. Vibrations in the fence are detected and analysed to determine whether they represent an intrusion attempt. The signals from the fence are processed to filter out as many false alarms as possible. It is important that the output from the sensors is linear along the fence, to ensure that there is no point on the perimeter less sensitive to intruder attack. The investigation reported in this paper aims to discover whether the performance of acoustic detection systems could be improved through research into fence acoustics. Using models of fence vibrations, sensors could be positioned to obtain a more linear response. In addition, through knowledge of the fence vibrations produced by both real attacks and the typical causes of false alarms, a system could be designed which could better distinguish between them, giving a lower false alarm rate and increased detection performance. Characteristic profiles were found for the different forms of attack to the fence. Although the investigation is still in its early stages, these attack profiles suggest that a new type of detection system could be designed based on profile recognition.