PrePass-Flow: A Machine Learning based technique to minimize ACL policy violation due to links failure in hybrid SDN

Wang, Lei; Muntean, Gabriel‐Miro; Akbar, Aamir; Shah, Nadir; Malik, Kaleem Razzaq

doi:10.1016/j.comnet.2020.107706

“…Additionally, the information provided is relative to each piece of equipment and with the 24h time frame, the predictions are very easily actionable. However, while the low definition of the failure hypothesis and the 24 hour window probably make the problem much easier to solve, it limits the utility of this method for larger and more Predict whether link between two network equipments will go down [34] Alarm Prediction Predict whether or not certain network states will lead to an alarm being raised (by an automated system or not) [49], [59] TABLE 3: Description of the different applications of NFP to networks heterogeneous networks as further diagnostics need to be run to find and solve the root cause of the failure. This paper also showcases the use of an alternative method of predicting future Key Performance Indicators (KPIs), and then detecting the fault in a second step.…”

Section: A Equipment Healthmentioning

confidence: 99%

“…Ibrar et al [34] work on a Hybrid SDN (HSDN) architecture composed of legacy (non SDN) and SDN activated switches. They identify the need in HSDN networks for legacy switches link failure information to be communicated quicker to the Controller.…”

Section: Link Failurementioning

confidence: 99%

Fault Prediction for Heterogeneous Networks using Machine Learning : a Survey

Murphy¹,

Lepers²,

Lavignotte³

2022

Preprint

1

0

View full text Add to dashboard Cite

<div>This manuscript provides a survey of Network Fault Prediction using Machine Learning from the years 1997 to 2021. Operational elements from the industry are shared to provide context. An overview of the different ML techniques that have been applied to Network Fault Prediction (NFP) is given. This is followed by a systematic review of the studies on NFP. Future research perspectives in the field are proposed.</div>

show abstract

“…We have already seen a system that goes in that direction in Ibrar et al [34], where the authors use an SDN controller to reroute traffic to avoid failing links. However if the objective is to completely neutralize the impact of failures, the system could be generalized to other types of networks (not only HSDN and not only switches), and alarms could be raised to ensure maintenance for faulty nodes.…”

Section: Perspectivesmentioning

confidence: 99%

Fault Prediction for Heterogeneous Networks using Machine Learning : a Survey

Murphy¹,

Lepers²,

Lavignotte³

2022

Preprint

1

0

View full text Add to dashboard Cite

<div>This manuscript provides a survey of Network Fault Prediction using Machine Learning from the years 1997 to 2021. Operational elements from the industry are shared to provide context. An overview of the different ML techniques that have been applied to Network Fault Prediction (NFP) is given. This is followed by a systematic review of the studies on NFP. Future research perspectives in the field are proposed.</div>

show abstract

“…Ibrar et al [12] present PrePass-Flow, a ML-based technique to minimize Access Control List (ACL) policy violation due to links failure in hybrid SDN. In hybrid SDN, an SDN controller can directly poll the status of the SDN links via the OpenFlow protocol.…”

Section: Machine Learning and Artificial Intelligencementioning

confidence: 99%