Intrusion Event Identification for Fiber Perimeter Security System Based on ARMA Modeling and Sigmoid Fitting

Huang Xiangdong; Wang Biyao; Liu Kun; Liu Tiegen

doi:10.3788/CJL202047.1006004

Chinese Journal of Lasers, Volume. 47, Issue 10, 1006004(2020)

Intrusion Event Identification for Fiber Perimeter Security System Based on ARMA Modeling and Sigmoid Fitting

Huang Xiangdong¹, Wang Biyao¹, Liu Kun², and Liu Tiegen²

Author Affiliations

¹School of Electrical and Information Engineering, Tianjin University, Tianjin 300072, China

²Department of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin 300072, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(10)

Fig. 1. Schematic diagram of DMZI disturbed optical fiber sensing system

Download full size

Fig. 2. Flowchart of intrusion event discrimination

Download full size

Fig. 3. Flowchart of the proposed intrusion event discrimination scheme

Download full size

Fig. 4. Sigmoid based fitness for occurrence probabilities of the knocking intrusion

Download full size

Fig. 5. Waveforms of 6 intrusion events. (a) Climbing; (b) knocking; (c) waggling; (d) cutting; (e) kicking; (f) crashing

Download full size

Fig. 6. Averaged feature vectors of 6 intrusion events. (a) Climbing; (b) knocking; (c) waggling; (d) cutting; (e) kicking; (f) crashing

Download full size

Table 1. Time consumption of two methods for intrusion event discrimination unit: s
View table
Table 1. Time consumption of two methods for intrusion event discrimination unit: s
Method EMD Proposed
Time consumption 13.274 1.057

Table 2. Pattern recognition rates of two methods for 4 intrusion events unit: %
View table
Table 2. Pattern recognition rates of two methods for 4 intrusion events unit: %
Method Climbing Knocking Waggling Cutting
EMD 87.3 70.9 99.7 85.1
Proposed 100 70 100 91.43

Table 3. Pattern recognition rates of two methods for 6 intrusion events unit: %
View table
Table 3. Pattern recognition rates of two methods for 6 intrusion events unit: %
Method Climbing Knocking Waggling Cutting Kicking Crashing
EMD 71.43 65.71 98.57 77.14 42.86 24.29
Proposed 100 75.71 97.14 91.43 85.71 72.86

Table 4. Estimation of occurrence probabilities for 6 intrusion events unit: %

View table

Table 4. Estimation of occurrence probabilities for 6 intrusion events unit: %

Realevent	Recognitionresult	Climbing	Knocking	Waggling	Cutting	Kicking	Crashing	Error rate (when two eventswith maximum probabilityvalues are taken into account)
Knocking	Cutting	0.79	44.23	0.61	52.80	0.38	0.19	1.97
Cutting	Knocking	0.73	62.76	0.75	35.18	0.39	0.19	2.05
Waggling	Crashing	1.35	3.10	33.56	2.98	2.50	56.52	9.93
Kicking	Crashing	0.68	1.22	1.27	0.73	36.22	59.88	3.90
Crashing	Kicking	0.56	1.07	1.22	0.6	53.36	43.19	3.45

Tools

Get Citation

Copy Citation Text

Huang Xiangdong, Wang Biyao, Liu Kun, Liu Tiegen. Intrusion Event Identification for Fiber Perimeter Security System Based on ARMA Modeling and Sigmoid Fitting[J]. Chinese Journal of Lasers, 2020, 47(10): 1006004

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Fiber optics and optical communication

Received: May. 6, 2020

Accepted: --

Published Online: Oct. 16, 2020

The Author Email:

DOI:10.3788/CJL202047.1006004

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Time consumption of two methods for intrusion event discrimination unit: s

Table 1. Time consumption of two methods for intrusion event discrimination unit: s

Table 2. Pattern recognition rates of two methods for 4 intrusion events unit: %

Table 2. Pattern recognition rates of two methods for 4 intrusion events unit: %

Table 3. Pattern recognition rates of two methods for 6 intrusion events unit: %

Table 3. Pattern recognition rates of two methods for 6 intrusion events unit: %

Table 4. Estimation of occurrence probabilities for 6 intrusion events unit: %

Table 4. Estimation of occurrence probabilities for 6 intrusion events unit: %