Name
Title | ||
|---|---|---|
Multi-Feature-Based Intrusion Detection for Optical Fiber Perimeter Security System: A Aase Study |
Abstract | ||
|---|---|---|
With the popularity of unattended stations, perimeter security technology has received close attention. Among many technologies, the optical fiber sensor based method is most widely used one due to its passive, low-cost, long-life, strong anti-interference ability and strong environmental adaptability. This paper proposes an intrusion detection method for optical fiber perimeter by combing multiple features with improved double-threshold method. First, three types of features are extracted, including short-time energy, short-time zero cross rate, and short-time wavelet coefficient energy. Then, an improved double-threshold method that constrained by the maximum length of silence and the minimum length of intrusion is proposed for detecting intrusions. Experiments on the optical fiber perimeter security system in Chongqing Fuling shale gas field show that the methods based on the short-time energy and the short-time wavelet coefficient energy can effectively detect the start points and end points of intrusions, and the performance of the short-time wavelet coefficient energy based method is the best. However, the detection accuracy of the short-time zero cross rate based method, is very low and cannot be applied independently. |
Year | DOI | Venue |
|---|---|---|
2018 | 10.1109/ICCI-CC.2018.8482013 | 2018 IEEE 17th International Conference on Cognitive Informatics & Cognitive Computing (ICCI*CC) |
Keywords | Field | DocType |
intrusion detection,short-time energy,short-time wavelet coefficient energy,double-threshold method,optical fiber perimeter security system,cognitive signal processing | Optical fiber,Zero crossing,Fiber optic sensor,Computer science,Feature extraction,Real-time computing,Perimeter,Vibration,Intrusion detection system,Wavelet | Conference |
ISBN | Citations | PageRank |
978-1-5386-3361-8 | 0 | 0.34 |
References | Authors | |
2 | 6 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Xiaohua Gu | 1 | 29 | 6.84 |
| Wang Tian | 2 | 17 | 15.16 |
| Song Hou | 3 | 0 | 0.34 |
| Jun Peng | 4 | 0 | 1.69 |
| Hongjin Wang | 5 | 0 | 0.68 |
| Qinfeng Xia | 6 | 0 | 1.01 |