Invention Grant
- Patent Title: Using machine learning to monitor link quality and predict link faults
-
Application No.: US15666015Application Date: 2017-08-01
-
Publication No.: US10298465B2Publication Date: 2019-05-21
- Inventor: Alam Yadav , Madhava N , Saikat Sanyal
- Applicant: Juniper Networks, Inc.
- Applicant Address: US CA Sunnyvale
- Assignee: Juniper Networks, Inc.
- Current Assignee: Juniper Networks, Inc.
- Current Assignee Address: US CA Sunnyvale
- Agency: Harrity & Harrity, LLP
- Main IPC: H04L12/24
- IPC: H04L12/24 ; H04L12/26 ; H04W24/04 ; H04W24/08

Abstract:
A device may receive a trained data model that has been trained using historical link quality information associated with a set of links. The device may determine, after receiving the trained data model, link quality information associated with a link that is actively supporting traffic. The device may classify the link by using the link quality information as input for the data model. The data model may classify the link into a class of a set of classes associated with measuring link quality. The device may determine an actual quality level of the link. The device may selectively update the class of the link after determining the actual link quality of the link. The device may perform one or more actions associated with improving link quality based on classifying the link and/or selectively updating the class of the link.
Public/Granted literature
- US20190044824A1 USING MACHINE LEARNING TO MONITOR LINK QUALITY AND PREDICT LINK FAULTS Public/Granted day:2019-02-07
Information query