Dual-Stage Soft Failure Detection and Identification for Low-Margin Elastic Optical Network by Exploiting Digital Spectrum Information

Shu, Liang; Yu, Zhenming; Wan, Zhiquan; Zhang, Jing; Hu, Shaohua; Xu, Kun

doi:10.1109/jlt.2019.2947562

Cited by 32 publications

(16 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The importance of knowledge about network topology for efficient failure detection and localization is emphasized in [16]. Failure detection and localization in optical networks is hot-topic area because these networks usually represent backbone and core of service provider communication network infrastructure [17][18][19][20][21][22]. Two types of failures can be observed, hard and soft failures.…”

Section: Related Workmentioning

confidence: 99%

“…Two types of failures can be observed, hard and soft failures. Hard failures represent complete malfunction of the device or link (for example, fiber cut), while the soft failures represent the degradation of performance (for example, component aging) [17]. Hard failures impact the network performance immediately and are easier to detect.…”

Section: Related Workmentioning

confidence: 99%

“…In our paper, the focus is on hard failures, but the big data platform and collected data can be used for soft failures detection as well; however, this aspect is not within the scope of this paper. Typically, machine learning and neural networks are used for soft failure detection in optical networks [17][18][19]. Similarly, machine learning can be used for failure localization in optical networks as well [20][21][22].…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Detection and Localization of Failures in Hybrid Fiber–Coaxial Network Using Big Data Platform

Simakovic

Cica

2021

Electronics

View full text Add to dashboard Cite

Modern HFC (Hybrid Fiber–Coaxial) networks comprise millions of users. It is of great importance for HFC network operators to provide high network access availability to their users. This requirement is becoming even more important given the increasing trend of remote working. Therefore, network failures need to be detected and localized as soon as possible. This is not an easy task given that there is a large number of devices in typical HFC networks. However, the large number of devices also enable HFC network operators to collect enormous amounts of data that can be used for various purposes. Thus, there is also a trend of introducing big data technologies in HFC networks to be able to efficiently cope with the huge amounts of data. In this paper, we propose a novel mechanism for efficient failure detection and localization in HFC networks using a big data platform. The proposed mechanism utilizes the already present big data platform and collected data to add one more feature to big data platform—efficient failure detection and localization. The proposed mechanism has been successfully deployed in a real HFC network that serves more than one million users.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Detection and Localization of Failures in Hybrid Fiber–Coaxial Network Using Big Data Platform

Simakovic

Cica

2021

Electronics

View full text Add to dashboard Cite

show abstract

“…As discussed in [2], several studies have demonstrated the potential of ML and data analytics in leveraging field data to automate ONFM and effectively perform Quality of Transmis-sion (QoT) monitoring [3], failure detection [4,5], failure-cause identification [6,7], failure localization [8][9][10][11] and failure prediction [12]. Most ML-based ONFM approaches rely on supervised learning techniques and on monitoring of signal-quality data, e.g., Optical Signal-to-Noise Ratio (OSNR) and/or Bit Error Rate (BER), made available by modern coherent receivers or by Optical Spectrum Analysers (OSAs).…”

Section: Introductionmentioning

confidence: 99%

“…In general, ML has been demonstrated to accurately recognize such failure signatures under specific assumptions (e.g., a single lightpath, a certain number of spans in an optical transmission system, etc.) [4,6]. However, a main limitation to practical ML deployments for ONFM comes from the fact that, in real network scenarios, availability of historical failure data can be scarce for several reasons (e.g., lack of monitoring equipment at every network node, high cost of acquisition of large datasets, high resilience of optical transmission systems, which makes failures occurrence rare phenomena, etc.).…”

Section: Introductionmentioning

confidence: 99%

Domain adaptation and transfer learning for failure detection and failure-cause identification in optical networks across different lightpaths [Invited]

Musumeci

Venkata

Hirota

et al. 2021

J. Opt. Commun. Netw.

View full text Add to dashboard Cite

Optical Network Failure Management (ONFM) is a promising application of Machine Learning (ML) to optical networking. Typical ML-based ONFM approaches exploit historical monitored data, retrieved in a specific domain (e.g., a link or a network), to train supervised ML models and learn failures characteristics (a signature) that will be helpful upon future failures occurrence in that domain. Unfortunately, in operational networks, data availability often constitutes a practical limitation to the deployment of MLbased ONFM solutions, due to scarce availability of labeled data comprehensively modeling all possible failure types. One could purposely inject failures to collect training data, but this is time consuming and not desirable by operators. A possible solution is Transfer Learning (TL), i.e., training ML models on a Source Domain (SD), e.g., a lab testbed, and then deploy trained models on a Target Domain (TD), e.g., an operator network, possibly fine-tuning the learned models by re-training with few TD data. Moreover, in those cases when TL re-training is not successful (e.g., due to intrinsic difference of SD and TD), another solution is domain adaptation, which consists of combining unlabeled SD and TD data before model training. We investigate domain adaptation and TL for failure detection and failure-cause identification across different lightpaths leveraging real OSNR data. We find that, for the considered scenarios, up to 20 percentage points of accuracy increase can be obtained with domain adaptation for failure detection, while for failure-cause identification only combining domain adaptation with model re-training provides significant benefit, reaching 4-5 percentage points of accuracy increase in the considered cases.

show abstract

A review of machine learning-based failure management in optical networks

Wang

Zhang²,

Chen³

et al. 2022

Sci. China Inf. Sci.

View full text Add to dashboard Cite

Failure management plays a significant role in optical networks. It ensures secure operation, mitigates potential risks, and executes proactive protection. Machine learning (ML) is considered to be an extremely powerful technique for performing comprehensive data analysis and complex network management and is widely utilized for failure management in optical networks to revolutionize the conventional manual methods. In this study, the background of failure management is introduced, where typical failure tasks, physical objects, ML algorithms, data sources, and extracted information are illustrated in detail. An overview of the applications of ML in failure management is provided in terms of alarm analysis, failure prediction, failure detection, failure localization, and failure identification. Finally, the future directions on ML for failure management are discussed from the perspective of data, model, task, and emerging techniques.

show abstract

Dual-Stage Soft Failure Detection and Identification for Low-Margin Elastic Optical Network by Exploiting Digital Spectrum Information

Cited by 32 publications

References 23 publications

Detection and Localization of Failures in Hybrid Fiber–Coaxial Network Using Big Data Platform

Detection and Localization of Failures in Hybrid Fiber–Coaxial Network Using Big Data Platform

Domain adaptation and transfer learning for failure detection and failure-cause identification in optical networks across different lightpaths [Invited]

A review of machine learning-based failure management in optical networks

Contact Info

Product

Resources

About