Wi-Fi access point performance management

• US 10,355,938 B2
• Filed: 06/13/2016
• Issued: 07/16/2019
• Est. Priority Date: 06/22/2015
• Status: Active Grant

• Current Assignee: Accenture Global Services Limited (Accenture PLC)
• Original Assignee: Accenture Global Services Limited (Accenture PLC)
• Inventors: Saha, Vivek, Chakraborty, Arnab, Sehgal, Sachin, Jain, Ankit, Bertrand, Eric, Kumar, Amit
• Primary Examiner(s): Lazaro, David R
• Assistant Examiner(s): Khurshid, Zia
• Application Number:
  US15/180,975
• Publication Number:
  US 20160373306A1
• Time in Patent Office: 1,128 Days
• Field of Search: 709223
• CPC Class Codes: H04L 41/06 Management of faults, events, alarms or notifications
• CPC Class Codes: H04L 41/12 Discovery or management of network topologies
• CPC Class Codes: H04L 41/147 for predicting network behaviour
• CPC Class Codes: H04W 24/08 Testing, supervising or monitoring using real traffic
• CPC Class Codes: H04W 4/80 Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication
• CPC Class Codes: H04W 84/12 WLAN [Wireless Local Area Networks]
• CPC Class Codes: H04W 88/08 Access point devices

Abstract

Wi-Fi access point performance management may include receiving a session analytic record related to a plurality of wireless access points, and analyzing the session analytic record related to the plurality of wireless access points to determine a root cause of at least one malfunctioning node related to at least one of the plurality of wireless access points, and predict failure of the at least one of the plurality of wireless access points. Further, a graphical user interface display or a report may be generated. The graphical user interface display or the report may be related to the determination of the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, and prediction of the failure of the at least one of the plurality of wireless access points.

Claims

1. A Wi-Fi access point performance management system comprising:

a data aggregator, executed by the at least one hardware processor, to

aggregate data from a plurality of data sources related to the at least one of a plurality of wireless access points to generate a session analytic record related to the plurality of wireless access points, wherein the data from the plurality of data sources includes device generated session data that represents Wi-Fi session records obtained from a plurality of wireless devices connected to the at least one of the plurality of wireless access points;

a performance monitor and predictor, executed by at least one hardware processor, to

analyze the session analytic record related to the plurality of wireless access points to

determine a root cause of at least one malfunctioning node related to the at least one of the plurality of wireless access points by recursive partitioning to generate a decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, wherein the root cause is based on a determination of a type of firmware installed on the at least one of the plurality of wireless access points,

predict failure of the at least one of the plurality of wireless access points by implementing logistic regression to determine an access point failure prediction model to predict failure of the at least one of the plurality of wireless access points by determining whether a percentage of sessions related to the at least one of the plurality of wireless access points are below a session quality metric, and in response to a determination that the percentage of sessions related to the at least one of the plurality of wireless access points are below the session quality metric, designating the at least one of the plurality of wireless access points as failed,

determine a problem that has been identified by the decision tree and the access point failure prediction model with respect to at least one of nodes related to the plurality of wireless access points, or the plurality of wireless access points, and

implement at least one corrective action to rectify the problem, wherein

the corrective action includes at least one of firmware upgrade, re-routing of traffic to another node, re-setting of a wireless access point, or re-setting of an associated component, and

when the corrective action is a firmware upgrade, a new decision tree is generated after the firmware upgrade;

an insight generator, executed by the at least one hardware processor, to generate at least one graphical user interface display or at least one report related to the

determination of the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, and

prediction of the failure of the at least one of the plurality of wireless access points; and

a non-transitory computer readable medium storing machine readable instructions that when executed by the at least one hardware processor cause the at least one hardware processor to;

control operations of a network including the at least one malfunctioning node and the at least one of the plurality of wireless access points based on the

determination of the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, and

prediction of the failure of the at least one of the plurality of wireless access points.

2. The Wi-Fi access point performance management system according to claim 1, wherein the at least one graphical user interface display or the at least one report related to the determination of the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points includes

the decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points.

3. The Wi-Fi access point performance management system according to claim 2, wherein the decision tree includes color coding to identify different latency ranges of the at least one malfunctioning node related to the at least one of the plurality of wireless access points.

4. The Wi-Fi access point performance management system according to claim 1, wherein the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points is based on at least one of

a determination of a number of events on a wireless LAN gateway related to the at least one of the plurality of wireless access points as a primary session,

a determination of a number of events on the wireless LAN gateway related to the at least one of the plurality of wireless access points as a secondary session, or a determination of whether a primary wireless LAN gateway is identical to a secondary wireless LAN gateway.

5. The Wi-Fi access point performance management system according to claim 1, wherein the at least one graphical user interface display or the at least one report related to the prediction of the failure of the at least one of the plurality of wireless access points includes

a plot of the at least one of the plurality of wireless access points relative to longitude and latitude information related to the at least one of the plurality of wireless access points, wherein the plot of the at least one of the plurality of wireless access points includes color coding to identify different types of predictions of the failure of the at least one of the plurality of wireless access points.

6. The Wi-Fi access point performance management system according to claim 1, wherein the at least one graphical user interface display related to the prediction of the failure of the at least one of the plurality of wireless access points includes

an option to select a performance status of the at least one of the plurality of wireless access points for a current time duration and a future time duration.

7. The Wi-Fi access point performance management system according to claim 1,

wherein the plurality of data sources further include at least one of

wireless LAN gateway data that represents gateway logs related to the at least one of the plurality of wireless access points, or

access point, node, and core health data that represents health check logs related to the at least one of the plurality of wireless access points.

8. The Wi-Fi access point performance management system according to claim 1, wherein the performance monitor and predictor is to determine the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points by

implementing recursive partitioning to determine a decision tree model to generate the decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points.

9. The Wi-Fi access point performance management system according to claim 8, wherein the performance monitor and predictor is to implementing recursive partitioning to determine the decision tree model to generate the decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points by

identifying, from a plurality of possible independent variables, a reduced set of independent variables that is significantly related to latency of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, and

using the identified reduced set of the independent variables to determine the decision tree model to generate the decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points.

10. The Wi-Fi access point performance management system according to claim 8, wherein the performance monitor and predictor is to implementing recursive partitioning to determine the decision tree model to generate the decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points by

identifying, from a plurality of possible independent variables, an independent variable that divides data from the session analytic record into two groups, and

applying data division to each group of the two groups until a predetermined condition is met.

11. The Wi-Fi access point performance management system according to claim 1, wherein the performance monitor and predictor is to predict failure of the at least one of the plurality of wireless access points by

implementing machine learning of a predetermined time duration of the session analytic record related to the plurality of wireless access points,

analyzing, based on the machine learning of the predetermined time duration of the session analytic record related to the plurality of wireless access points, a further predetermined time duration of the session analytic record related to the plurality of wireless access points, and

predicting, based on the analysis of the further predetermined time duration of the session analytic record related to the plurality of wireless access points, failure of the at least one of the plurality of wireless access points for a future predetermined time duration associated with the further predetermined time duration.

12. The Wi-Fi access point performance management system according to claim 11, wherein the future predetermined time duration represents a time duration for which there is no available session analytic record related to the plurality of wireless access points.

13. The Wi-Fi access point performance management system according to claim 1, wherein the performance monitor and predictor is to analyze the session analytic record related to the plurality of wireless access points to predict failure of the at least one of the plurality of wireless access points by

determining whether a percentage of sessions related to the at least one of the plurality of wireless access points are below a session quality metric, and

in response to a determination that the percentage of sessions related to the at least one of the plurality of wireless access points are below the session quality metric, designating the at least one of the plurality of wireless access points as failed.

14. The Wi-Fi access point performance management system according to claim 1, wherein the performance monitor and predictor is to analyze the session analytic record related to the plurality of wireless access points to predict failure of the at least one of the plurality of wireless access points by

analyzing a relationship of a dependent variable related to a failed wireless access point to a plurality of independent variables related to the at least one of the plurality of wireless access points, and

iteratively determining, from the plurality of independent variables, a set of independent variables that accurately maps to the dependent variable related to the failed wireless access point.

15. The Wi-Fi access point performance management system according to claim 1, further comprising:

a model deployer, executed by the at least one hardware processor, to

track, based on a second decision tree, a result of a modification related to an attribute of the at least one of the plurality of wireless access points, by comparing the second decision tree to a first decision tree that represents the at least one malfunctioning node related to the at least one of the plurality of wireless access points prior to the modification related to the attribute of the at least one of the plurality of wireless access points.

16. A method for Wi-Fi access point performance management, the method comprising:

aggregating, by at least one hardware processor, data from a plurality of data sources related to at least one of a plurality of wireless access points to generate a session analytic record related to the plurality of wireless access points, wherein the data from the plurality of data sources includes device generated session data that represents Wi-Fi session records obtained from a plurality of wireless devices connected to the at least one of the plurality of wireless access points;

analyzing, by the at least one hardware processor, the session analytic record related to the plurality of wireless access points to determine a root cause of at least one malfunctioning node related to the at least one of the plurality of wireless access points by recursive partitioning to generate a decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, wherein the root cause is based on a determination of a type of firmware installed on the at least one of the plurality of wireless access points;

predicting, by the at least one hardware processor, failure of the at least one of the plurality of wireless access points by implementing logistic regression to determine an access point failure prediction model to predict failure of the at least one of the plurality of wireless access points by determining whether a percentage of sessions related to the at least one of the plurality of wireless access points are below a session quality metric, and in response to a determination that the percentage of sessions related to the at least one of the plurality of wireless access points are below the session quality metric, designating the at least one of the plurality of wireless access points as failed;

determining, by the at least one hardware processor, a problem that has been identified by the decision tree and the access point failure prediction model with respect to at least one of nodes related to the plurality of wireless access points, or the plurality of wireless access points;

implementing, by the at least one hardware processor, at least one corrective action to rectify the problem, wherein

the corrective action includes at least one of firmware upgrade, re-routing of traffic to another node, re-setting of a wireless access point, or re-setting of an associated component, and

when the corrective action is a firmware upgrade, a new decision tree is generated after the firmware upgrade; and

generating, by the at least one hardware processor, at least one graphical user interface display or at least one report related to the determination of the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, and related to the prediction of the failure of the at least one of the plurality of wireless access points.

17. The method for Wi-Fi access point performance management according to claim 16, wherein analyzing the session analytic record related to the plurality of wireless access points to determine the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points further comprises at least one of:

determining a type of operating system related to the at least one of the plurality of wireless access points;

determining an account status related to the at least one of the plurality of wireless access points;

or

determining a combined rating of a wireless LAN gateway related to the at least one of the plurality of wireless access points.

18. A non-transitory computer readable medium having stored thereon machine readable instructions for Wi-Fi access point performance management, the machine readable instructions when executed cause a computer system to:

aggregate data from a plurality of data sources related to at least one of a plurality of wireless access points to generate a session analytic record related to the plurality of wireless access points, wherein the data from the plurality of data sources includes device generated session data that represents Wi-Fi session records obtained from a plurality of wireless devices connected to the at least one of the plurality of wireless access points;

analyze the session analytic record related to the plurality of wireless access points to determine a root cause of at least one malfunctioning node related to the at least one of the plurality of wireless access points by recursive partitioning to generate a decision tree that identifies the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, wherein the root cause is based on a determination of a type of firmware installed on the at least one of the plurality of wireless access points;

predict failure of the at least one of the plurality of wireless access points by implementing logistic regression to determine an access point failure prediction model to predict failure of the at least one of the plurality of wireless access points by determining whether a percentage of sessions related to the at least one of the plurality of wireless access points are below a session quality metric, and in response to a determination that the percentage of sessions related to the at least one of the plurality of wireless access points are below the session quality metric, designating the at least one of the plurality of wireless access points as failed;

determine a problem that has been identified by the decision tree and the access point failure prediction model with respect to at least one of nodes related to the plurality of wireless access points, or the plurality of wireless access points;

implement at least one corrective action to rectify the problem, wherein

the corrective action includes at least one of firmware upgrade, re-routing of traffic to another node, re-setting of a wireless access point, or re-setting of an associated component, and

when the corrective action is a firmware upgrade, a new decision tree is generated after the firmware upgrade;

generate at least one graphical user interface display or at least one report related to the determination of the root cause of the at least one malfunctioning node related to the at least one of the plurality of wireless access points, and related to the prediction of the failure of the at least one of the plurality of wireless access points by plotting the at least one of the plurality of wireless access points relative to longitude and latitude information related to the at least one of the plurality of wireless access points; and

code the plot of the at least one of the plurality of wireless access points to identify different types of predictions of the failure of the at least one of the plurality of wireless access points.

19. The non-transitory computer readable medium according to claim 18, wherein the machine readable instructions to analyze the session analytic record related to the plurality of wireless access points, further comprise machine readable instructions to:

implement machine learning of a predetermined time duration of the session analytic record related to the plurality of wireless access points;

analyze, based on the machine learning of the predetermined time duration of the session analytic record related to the plurality of wireless access points, a further predetermined time duration of the session analytic record related to the plurality of wireless access points; and

predict, based on the analysis of the further predetermined time duration of the session analytic record related to the plurality of wireless access points, failure of the at least one of the plurality of wireless access points for a future predetermined time duration associated with the further predetermined time duration.