Event Discrimination Method for Distributed Optical Fiber Intrusion Sensing System Based on Integrated Time/Frequency Domain Feature Extraction

Kuan Peng; Cheng Feng; Senmao Wang; Fan Ai; Hao Li; Deming Liu; Qizhen Sun

doi:10.3788/AOS201939.0628002

Acta Optica Sinica, Volume. 39, Issue 6, 0628002(2019)

Event Discrimination Method for Distributed Optical Fiber Intrusion Sensing System Based on Integrated Time/Frequency Domain Feature Extraction

Kuan Peng^1,2, Cheng Feng¹, Senmao Wang¹, Fan Ai¹, Hao Li¹, Deming Liu¹, and Qizhen Sun^1、*

Author Affiliations

¹ School of Optical and Electronic Information, Huazhong University of Science and Technology, Wuhan, Hubei 430074, China

² Wuhan Fisilink Microelectronics Technology Company Limited, Wuhan, Hubei 430074, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(15)

Fig. 1. Principle diagram of WLI-DOFVS system

Download full size

Fig. 2. Flow chart of data processing

Download full size

Fig. 3. Schematic of Mallat multiresolution decomposition

Download full size

Fig. 4. Signals before and after denoising by wavelet threshold. (a) Original signal; (b) denoised signal

Download full size

Fig. 5. Signal segmentation results of several typical intrusion events. (a) Footsteps of passerby; (b) bicycle; (c) knocking on fence; (d) cutting of optical cable

Download full size

Fig. 6. Photograph of sensing fiber laying on fence and ground

Download full size

Fig. 7. Typical features in time domain. (a) Average fragment interval; (b)fragment length; (c) PAR

Download full size

Fig. 8. Energy distributions in frequency domain. (a) Footsteps of passerby; (b) bicycle rolling; (c) knocking on the fence; (d) cutting of optical cable

Download full size

Table 1. Main parameters of WLI-DOFVS system
View table
Table 1. Main parameters of WLI-DOFVS system
Parameter Value
${L_{0}}_{}$ /km 50
${L_{d}}_{}$ /km 1
W /nm ≈40
P /dBm 9.4

Table 2. Recognition rate of 4 different methods for the 1st training set%

View table

Table 2. Recognition rate of 4 different methods for the 1st training set%

Intrusion event	Time domain feature +SVM	Frequency domain feature +SVM	Time/frequency domain feature +RBF NN	Time/frequency domain feature +SVM
Footsteps of passerby	100	52.5	47.5	97.5
Bicycle rolling	75.0	97.5	100	100
Knocking on fence	90.0	97.5	87.5	100
Cutting of optical cable	100	72.5	87.5	95.0
Average recognition rate	91.25	80.00	80.63	98.13

Table 3. Recognition rate of 4 different methods for the 2nd training set%

View table

Table 3. Recognition rate of 4 different methods for the 2nd training set%

Intrusion event	Time domain feature +SVM	Frequency domain feature +SVM	Time/frequency domain feature +RBF NN	Time/frequency domain feature +SVM
Footsteps of passerby	100	50.0	82.5	98.75
Bicycle rolling	90.0	95.0	97.5	95.0
Knocking on fence	85.0	95.0	87.5	100
Cutting of optical cable	100	60.0	72.5	100
Average recognition rate	93.75	75.00	85.00	98.50

Table 4. Recognition rate of 4 different methods for the 3rd training set%

View table

Table 4. Recognition rate of 4 different methods for the 3rd training set%

Intrusion event	Time domain feature +SVM	Frequency domain feature +SVM	Time/frequency domain feature +RBF NN	Time/frequency domain feature +SVM
Footsteps of passerby	100	52.5	52.5	100
Bicycle rolling	92.5	97.5	92.5	97.5
Knocking on fence	87.5	97.5	97.5	100
Cutting of optical cable	97.5	70.0	87.5	100
Average recognition rate	94.38	79.38	82.50	99.38

Table 5. Recognition rate of 4 different methods for the 4th training set%

View table

Table 5. Recognition rate of 4 different methods for the 4th training set%

Intrusion event	Time domain feature +SVM	Frequency domain feature +SVM	Time/frequency domain feature +RBF NN	Time/frequency domain feature +SVM
Footsteps of passerby	100	40.0	40.0	100
Bicycle rolling	85.0	90.0	95.0	97.5
Knocking on fence	85.0	97.5	95.0	100
Cutting of optical cable	100	65.0	55.0	100
Average recognition rate	92.50	73.13	71.25	99.38

Table 6. Recognition rate of 4 different methods for the testing set%

View table

Table 6. Recognition rate of 4 different methods for the testing set%

Intrusion event	Time domain feature +SVM	Frequency domain feature +SVM	Time/frequency domain feature +RBF NN	Time/frequency domain feature +SVM
Footsteps of passerby	100	95.0	30.0	97.5
Bicycle rolling	92.5	97.5	95.0	100
Knocking on fence	75.0	90.0	82.5	92.5
Cutting of optical cable	97.5	2.5	42.5	95.0
Average recognition rate	91.25	71.50	62.50	96.25

Table 7. Mean and variance of recognition rate of 4 different methods for the 5 groups of data

View table

Table 7. Mean and variance of recognition rate of 4 different methods for the 5 groups of data

Recognition method	Mean /%	Variance /10^-4
Time domain feature +SVM	92.63	2.04
Frequency domain feature +SVM	75.80	14.18
Time/frequency domain feature +RBF NN	76.38	87.20
Time/frequency domain feature +SVM	98.33	1.65

Tools

Get Citation

Copy Citation Text

Kuan Peng, Cheng Feng, Senmao Wang, Fan Ai, Hao Li, Deming Liu, Qizhen Sun. Event Discrimination Method for Distributed Optical Fiber Intrusion Sensing System Based on Integrated Time/Frequency Domain Feature Extraction[J]. Acta Optica Sinica, 2019, 39(6): 0628002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites