Quality of Transmission Prediction with Machine Learning for Dynamic Operation of Optical WDM Networks

Samadi, Payman; Amar, Djamel; Lepers, Catherine; Lourdiane, Mounia; Bergman, Keren

doi:10.1109/ecoc.2017.8346216

Cited by 29 publications

(12 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Chen et al [27] proposed a framework for a knowledge-based autonomous service provisioning which is enabled by a deep neural network-based traffic estimator for multi-domain software-defined elastic optical networks. Samadi et al [28] proposed a neural network-based cognitive scalable method for dynamic provisioning of optical physical resources without need of prior network specific knowledge. Meng et al [29] provided a self-learning network that uses real-time monitoring together with Markov Chain Monte Carlo simulations.…”

Section: A Machine Learning Techniques In Communication Networkmentioning

confidence: 99%

Naïve Bayes Classifier-Assisted Least Loaded Routing for Circuit-Switched Networks

Zhang

Chen³

et al. 2019

IEEE Access

View full text Add to dashboard Cite

Circuit switching is a de facto switching technology widely employed in today's networks where the conventional approaches to routing have remained unchanged for many years. This paper develops a new and very different methodology, by incorporating a supervised naïve Bayes (NB) classifier, to assist least loaded (LL) routing and to further improve its performance that has remained the best among all the routing approach for the past several decades. Specifically, by iteratively learning the information of historical network snapshots, the NB classifier predicts potential future circuit blocking probability between each node pair if a service connection is established via a certain route between the node pair. The snapshots are taken for each service request arriving at an operating network that keeps on accepting and releasing dynamic service connections and records the number of busy capacity units on each link at each snapshot instance. The candidate route for serving a new request is selected based on both link loads and the potential future blocking probability in the entire network in case that this route is indeed used. The performance of the proposed approach is studied via simulations and compared with the conventional LL algorithm. The results indicate that the supervised NB classifier-assisted LL routing algorithm effectively reduces the blocking probability of service connections and outperforms the conventional LL routing algorithm. To speed up the learning process (which is based on a large number of network snapshots), we also develop a framework to incorporate the proposed approach in a parallel learning system. A network control system supporting online NB classifier-assisted LL routing algorithm is also described. INDEX TERMS Machine learning, naïve Bayes classifier, least loaded routing, blocking probability, circuit switched network.

show abstract

Section: A Machine Learning Techniques In Communication Networkmentioning

confidence: 99%

Naïve Bayes Classifier-Assisted Least Loaded Routing for Circuit-Switched Networks

Zhang

Chen³

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The considered models are: K-nearest neighbor, logistic regression, support vector machines, and artificial neural networks. In [145], the authors realized QoT prediction using machine learning for dynamic operation of optical WDM networks. The AI-assisted failure localization and anomaly detection can quickly and accurately identify the number of failures and the location of each failure.…”

Section: Ai-assisted Sensing and Survivability Techniquesmentioning

confidence: 99%

Survivable Deployments of Optical Sensor Networks against Multiple Failures and Disasters: A Survey

Zhang

Xin

2019

Sensors

View full text Add to dashboard Cite

Optical sensing that integrates communication and sensing functions is playing a more and more important role in both military and civil applications. Incorporating optical sensing and optical communication, optical sensor networks (OSNs) that undertake the task of high-speed and large-capacity applications and sensing data transmissions have become an important communication infrastructure. However, multiple failures and disasters in OSNs can cause serious sensing provisioning problems. To ensure uninterrupted sensing data transmission, survivability has always been an important research emphasis. This paper focuses on the survivable deployment of OSNs against multiple failures and disasters. We first review and evaluate the existing survivability technologies developed for or applied to OSNs, such as fiber bus protection, self-healing architecture, and 1 + 1 protection. We then elaborate on the disaster-resilient survivability requirement of OSNs. Moreover, we propose a new k-node (edge) sensing connectivity concept, which ensures the connectivity between sensing data and users. Based on k-node (edge) sensing connectivity, the disaster-resilient survivability technologies are developed. The key technologies necessary to implement k-node (edge) sensing connectivity are also elaborated. Recently, artificial intelligence (AI) has developed rapidly. It can be used to improve the survivability of OSNs. This paper details potential development directions of survivability technologies of optical sensing in OSNs employing AI.

show abstract

“…The estimation of QoT before establishing the lightpath in dynamic and agile network is necessary. Due to the conventional OSNR analytical models estimate QoT showing inefficient performances, such as requiring component specifications of network, over provisioning with increasing margins and resource under-utilization, dependent to the accuracy of the modeling, and inaccurate model, it is necessary to propose a novel cognitive method to achieve more optimal OSNR estimation [78]. Neural network with the ML engine could address this problem with no prior network knowledge requirements, which is scalable to large size networks, high accuracy, and adjusting the dynamic condition of network.…”

Section: Optical System Condition Monitoringmentioning

confidence: 99%

Towards converged, collaborative and co-automatic (3C) optical networks

Ji¹,

Zhang²,

Wang³

et al. 2018

Sci. China Inf. Sci.

View full text Add to dashboard Cite

The interconnection of all things is developing a new diagram of future information networks. However, it is difficult to realize future applications with only one single technique. Collaboration between multiple advanced techniques is leading the way for the development of future information networks. Optical communication is an enabling technique to achieve high speed, long reach, and low latency communication, which plays an important role on the transformation of information networks. To achieve these advantages that caters to the characteristics of future information networks, collaboration of multiple advanced techniques with optical, which is called "optical plus X", could realize the vision of "all things connected with networks". In this paper, we focus on the collaboration between optical networks with other techniques, mainly discuss four representative aspects, which are "optical plus IP", "optical plus radio", "optical plus computing", and "optical plus AI". We discuss the challenges, timely works, and developing trends. Finally, we give the future visions for optical network towards a collaborative, converged and co-automatic optical network. Keywords 5G optical transport networks, edge computing optical networks, IP over optical networks, AI-assisted optical networks, 3C optical networks

show abstract

Quality of Transmission Prediction with Machine Learning for Dynamic Operation of Optical WDM Networks

Cited by 29 publications

References 9 publications

Naïve Bayes Classifier-Assisted Least Loaded Routing for Circuit-Switched Networks

Naïve Bayes Classifier-Assisted Least Loaded Routing for Circuit-Switched Networks

Survivable Deployments of Optical Sensor Networks against Multiple Failures and Disasters: A Survey

Towards converged, collaborative and co-automatic (3C) optical networks

Contact Info

Product

Resources

About