We propose a novel approach to the recognition of particular classes of non-conventional events in signals from phase-sensitive optical time-domain-reflectometry-based sensors. Our algorithmic solution has two main features: filtering aimed at the de-nosing of signals and a Gaussian mixture model to cluster them. We test the proposed algorithm using experimentally measured signals. The results show that two classes of events can be distinguished with the best-case recognition probability close to 0.9 at sufficient numbers of training samples.

1.
For a review, see
X.
Bao
and
L.
Chen
,
Sensors
12
,
8601
(
2012
).
2.
Y.
Shi
,
H.
Feng
, and
Z.
Zeng
,
Sensors
15
,
21957
(
2015
).
3.
A. K.
Fedorov
,
V. A.
Lazarev
,
I. P.
Makhrov
,
N. O.
Pozhar
,
M. N.
Anufriev
,
A. B.
Pniov
, and
V. E.
Karasik
,
J. Phys.: Conf. Ser.
594
,
012049
(
2015
).
4.
H. F.
Taylor
and
C. E.
Lee
, U.S. patent 5, 194847 (16 March
1993
).
5.
J.
Park
,
W.
Lee
, and
H. F.
Taylor
,
Proc. SPIE
3555
,
49
(
1998
).
6.
K. N.
Choi
and
H. F.
Taylor
,
IEEE Photonics Technol. Lett.
15
,
386
(
2003
).
7.
Yu. N.
Kulchin
,
O. B.
Vitrik
,
A. V.
Dyshlyuk
,
A. M.
Shalagin
,
S. A.
Babin
, and
A. A.
Vlasov
,
Laser Phys.
17
,
1335
(
2007
).
8.
M. P.
Gold
,
J. Lightwave Technol.
3
,
39
(
1985
).
9.
J. C.
Juarez
,
E. W.
Maier
,
K. N.
Choi
, and
H. F.
Taylor
,
J. Lightwave Technol.
23
,
2081
(
2005
);
J. C.
Juarez
and
H. F.
Taylor
,
Opt. Lett.
30
,
3284
(
2005
).
[PubMed]
10.
Z.
Zhang
and
X.
Bao
,
Opt. Express
16
,
10240
(
2008
).
11.
Y. J.
Rao
,
J.
Luo
,
Z. L.
Ran
,
J. F.
Yue
,
X. D.
Luo
, and
Z.
Zhou
,
Proc. SPIE
7503
,
75031O
(
2009
).
12.
Y.
Lu
,
T.
Zhu
,
L.
Chen
, and
X.
Bao
,
J. Lightwave Technol.
28
,
3243
(
2010
).
13.
Z.
Qin
,
T.
Zhu
, and
X.
Bao
,
IEEE Photonics Technol. Lett.
23
,
1091
(
2011
).
14.
H. F.
Martins
,
S.
Martin-Lopez
,
P.
Corredera
,
P.
Salgado
,
O.
Frazão
, and
M.
González-Herráez
,
Opt. Lett.
38
,
872
(
2013
);
[PubMed]
H. F.
Martins
,
S.
Martin-Lopez
,
P.
Corredera
,
M. L.
Filograno
,
O.
Frazão
, and
M.
González-Herráez
,
J. Lightwave Technol.
31
,
3631
(
2013
).
15.
Z.
Qin
,
L.
Chen
, and
X.
Bao
,
IEEE Photonics Technol. Lett.
24
,
542
(
2012
).
16.
Y.
Shi
,
H.
Feng
,
Y.
An
,
X.
Feng
, and
Z.
Zeng
, in
Proceedings of the IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM)
(
IEEE
,
2014
), p.
1177
.
17.
N.
Duan
,
F.
Peng
,
Y.-J.
Rao
,
J.
Du
, and
Y.
Lin
,
SPIE Proc.
9157
,
91577A
(
2014
).
18.
Q.
Li
,
C.
Zhang
,
L.
Li
, and
X.
Zhong
,
Optik
125
,
2099
2103
(
2014
).
19.
A. K.
Fedorov
,
M. N.
Anufriev
,
A. A.
Zhirnov
,
E. T.
Nesterov
,
D. E.
Namiot
,
A. B.
Pnev
, and
V. E.
Karasik
,
Int. J. Open Inform. Technol.
3
,
16
(
2015
).
20.
W. B.
Lyons
and
E.
Lewis
,
Trans. Inst. Meas. Control
22
,
385
(
2000
).
21.
Q.
Sun
,
H.
Feng
,
X.
Yan
, and
Z.
Zeng
,
Sensors
15
,
15179
(
2015
).
22.
B.
Wang
,
S.
Pi
,
Q.
Sun
, and
B.
Jia
,
Opt. Eng.
54
,
055104
(
2015
).
23.
H.
Wu
,
S.
Xiao
,
X.
Li
,
Z.
Wang
,
J.
Xu
, and
Y.
Rao
,
J. Lightwave Technol.
33
,
3156
(
2015
).
24.

We use a semiconductor laser with an erbium-doped fiber amplifier with power limited to 1 W. The wavelength of the probe signal was 1550 nm, duration of the probe pulse was 100–500 ns, coherence length of the laser is 30 km and the signal was launched into a standard single-mode telecommunication optical fiber SMF28 of length approximately 50 km. The fiber is probed on ranges of a few km. The use of multimode fibers is not needed for our purposes because it affects spatial resolution.

25.
A. B.
Pnev
,
A. A.
Zhirnov
,
K. V.
Stepanov
,
E. T.
Nesterov
,
D. A.
Shelestov
, and
V. E.
Karasik
,
J. Phys.: Conf. Ser.
584
,
012016
(
2015
).
26.
E. T.
Nesterov
,
A. A.
Zhirnov
,
K. V.
Stepanov
,
A. B.
Pnev
,
V. E.
Karasik
,
Ya. A.
Tezadov
,
E. V.
Kondrashin
, and
A. B.
Ushakov
,
J. Phys.: Conf. Ser.
584
,
012028
(
2015
).
27.
T.
Huang
,
G.
Yang
, and
G.
Tand
,
IEEE Trans. Acoust., Speech, Signal Process.
27
,
13
(
1979
).
28.
X.
Huang
,
A.
Acero
, and
H.-W.
Hon
,
Spoken Language Processing: A Guide to Theory, Algorithm, and System Development
(
Prentice Hall
,
2001
).
29.
D.
O’Shaughnessy
,
Speech Communication: Human and Machine
(
Addison-Wesley
,
1987
).
30.
D.
Yu
and
D.
Li
, in
Automatic Speech Recognition
(
Springer-Verlag London
,
2015
), Chap. 2.
31.
T.
Hastie
,
R.
Tibshirani
, and
J.
Friedman
,
The Elements of Statistical Learning
(
Springer
,
New York
,
2001
).
33.
Here, SNR is defined as the ratio of the change in the signal with an event to the change in the signal without any event (i.e., changes relating to a hindrance in the environment and technical noise of the system).
34.

Since the recognition algorithm assigns to the class input events only to existed classes, it is important to exclude the effect of different types, which is beyond the classes considered here. For this, one needs to define through a preliminary test a probability threshold pc. If the algorithm predicts that an input event belongs to the first class with probability p1, this event belongs to the second type with probability p2 = 1 − p1. The workflow of the algorithm stops in cases where one of these probabilities is less than the threshold.

You do not currently have access to this content.