Leveraging Statistical Machine Learning to Address Failure Localization in Optical Networks

T. Panayiotou; S. P. Chatzis; G. Ellinas

doi:10.1364/JOCN.10.000162

Journal of Optical Communications and Networking
Vol. 10,
Issue 3,
pp. 162-173
(2018)
•https://doi.org/10.1364/JOCN.10.000162

Leveraging Statistical Machine Learning to Address Failure Localization in Optical Networks

T. Panayiotou, S. P. Chatzis, and G. Ellinas

Not Accessible

Your library or personal account may give you access

Get PDF
Email
Share
Get Citation
Copy Citation Text
T. Panayiotou, S. P. Chatzis, and G. Ellinas, "Leveraging Statistical Machine Learning to Address Failure Localization in Optical Networks," J. Opt. Commun. Netw. 10, 162-173 (2018)

Export Citation
- BibTex
- Endnote (RIS)
- HTML
- Plain Text
Citation alert
Save article

Check for updates

Abstract

In this work we consider the problem of fault localization in transparent optical networks. We attempt to localize single-link failures by utilizing statistical machine learning techniques trained on data that describe the network state upon current and past failure incidents. In particular, a Gaussian process classifier is trained on historical data extracted from the examined network, with the goal of modeling and predicting the failure probability of each link therein. To limit the set of suspect links for every failure incident, the proposed approach is complemented by the utilization of a graph-based correlation heuristic. The proposed approach is tested on a number of datasets generated for an orthogonal frequency-division multiplexing-based optical network, and demonstrates that the approach achieves a high localization accuracy (91%–99%) that is insignificantly affected as the size of the historical dataset is reduced. The approach is also compared to a conventional fault localization method that is based on the utilization of monitoring information. It is shown that the conventional method significantly increases the network cost, as measured by the number of monitoring nodes required to achieve the same accuracy as that achieved by the proposed approach. The proposed scheme can be used by service providers to reduce the network cost related to the fault localization procedure. As the approach is generic and does not depend on specific network technologies, it can be applied to different network types, e.g., fixed-grid or space-division multiplexing elastic optical networks.

Full Article | PDF Article

More Like This

Suspect fault screen assisted graph aggregation network for intra-/inter-node failure localization in ROADM-based optical networks

Ruikun Wang, Jiawei Zhang, Shuangyi Yan, Chuidian Zeng, Hao Yu, Zhiqun Gu, Bojun Zhang, Tarik Taleb, and Yuefeng Ji
J. Opt. Commun. Netw. 15(7) C88-C99 (2023)

Machine learning models for alarm classification and failure localization in optical transport networks

Jatin Babbar, Ahmed Triki, Reda Ayassi, and Maxime Laye
J. Opt. Commun. Netw. 14(8) 621-628 (2022)

Machine learning framework for timely soft-failure detection and localization in elastic optical networks

Sadananda Behera, Tania Panayiotou, and Georgios Ellinas
J. Opt. Commun. Netw. 15(10) E74-E85 (2023)

Previous Article Next Article

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Figures (6)

You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Tables (14)

You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Equations (14)

You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Link	Length (km)	$λ_{j}$	$β_{j}$
$e_{1}$	1100	471	2.77
$e_{2}$	1600	202	1.55
$e_{3}$	2800	131	1.45
$e_{4}$	600	971	2.53
$e_{5}$	1100	321	1.62
$e_{6}$	2000	67	2.71
$e_{7}$	600	1454	2.1
$e_{8}$	800	229	1.77
$e_{9}$	800	629	2.22
$e_{10}$	1200	674	2.65
$e_{11}$	700	263	2.43
$e_{12}$	900	636	2.5
$e_{13}$	700	1094	2.1

Traffic load (Erlangs)	7	10	20
# of blocked requests	0	0	0
Max. # of established lightpaths	20	23	38
Avg. # of established lightpaths	4.2	5.41	10.58
Min. # of established lightpaths	1	1	1

Link	7 Erlangs	10 Erlangs	20 Erlangs
$e_{1}$	771	550	299
$e_{2}$	1937	1568	1388
$e_{3}$	1819	1697	1542
$e_{4}$	1102	762	424
$e_{5}$	743	570	410
$e_{6}$	3807	3685	3570
$e_{7}$	444	242	61
$e_{8}$	927	774	608
$e_{9}$	546	313	135
$e_{10}$	686	467	153
$e_{11}$	812	472	202
$e_{12}$	632	415	58
$e_{13}$	694	481	238

Traffic load (Erlangs)	7	10	20
# of incidents	1541	1184	686
Min. # of suspect links	2	2	2
Max. # of suspect links	12	9	5
Avg. # of suspect links	3.24	2.77	2.18

Traffic load (Erlangs)	7	10	20
# of incidents in $D^{test}$	3000	3000	3000
# of correctly classified incidents by GBC	1459	1816	2314
# of incidents in $D_{r}^{test}$ (passed to GP)	1541	1184	686
# of correctly classified incidents by GP	1327	1068	655
GP accuracy	0.86	0.9	0.95
Total accuracy (GBC and GP)	0.93	0.96	0.99

Link	7 Erlangs	10 Erlangs	20 Erlangs
$e_{1}$	536	387	182
$e_{2}$	1361	1155	1035
$e_{3}$	1294	1223	1104
$e_{4}$	748	578	242
$e_{5}$	529	412	286
$e_{6}$	2694	2621	2538
$e_{7}$	330	176	39
$e_{8}$	645	573	438
$e_{9}$	394	211	97
$e_{10}$	501	330	102
$e_{11}$	543	322	139
$e_{12}$	456	290	42
$e_{13}$	491	356	179

# of incidents in $D^{train}$	7000	5000	3000	1000
Training time (min)	60	30	10	2
Classification time (s)	1.3	0.7	0.7	0.1

Link	7 Erlangs	10 Erlangs	20 Erlangs
$e_{1}$	337	240	120
$e_{2}$	795	709	593
$e_{3}$	810	761	686
$e_{4}$	449	330	162
$e_{5}$	330	250	176
$e_{6}$	1614	1571	1519
$e_{7}$	186	97	26
$e_{8}$	371	328	256
$e_{9}$	233	105	61
$e_{10}$	291	198	54
$e_{11}$	326	194	83
$e_{12}$	274	189	21
$e_{13}$	308	222	107

Link	7 Erlangs	10 Erlangs	20 Erlangs
$e_{1}$	337	240	120
$e_{2}$	795	709	593
$e_{3}$	810	761	686
$e_{4}$	449	330	162
$e_{5}$	330	250	176
$e_{6}$	1614	1571	1519
$e_{7}$	186	97	26
$e_{8}$	371	328	256
$e_{9}$	233	105	61
$e_{10}$	291	198	54
$e_{11}$	326	194	83
$e_{12}$	274	189	21
$e_{13}$	308	222	107

Leveraging Statistical Machine Learning to Address Failure Localization in Optical Networks

Author Affiliations

Abstract

Cited By

Figures (6)

Tables (14)

Equations (14)

Journal of Optical Communications and Networking