Granted Patent B2

US 12,231,285 · App. 18/268,011 · Granted Feb 18, 2025

Inventors: Pino Castrogiovanni (Turin, IT); Roberta Giannantonio (Turin, IT); Ennio Grasso (Turin, IT); Pier Giorgio Santandrea (Bologna, IT)

Assignee: Telecom Italia S.p.A.

H04L41/0604H04L41/147H04L41/16H04L41/5074

Abstract

Claims (75)

1. A system, comprising a plurality of hardware network nodes, including a processor with non-transitory computer readable media, for managing network alarms generated by element manager modules in response to corresponding faults/malfunctioning affecting network nodes or links of a telecommunication network, wherein the system stores software that, when executed by the system, causes the system to:

gather network alarms and generate a first set of network alarms comprising at least a subset of the gathered network alarms;

arrange the network alarms of the first set into a network alarm table providing, for each network alarm listed in the network alarm table, data describing the network alarm, wherein the system manages faults/malfunctioning affecting network nodes or links corresponding to the network alarms based on the network alarm table; and

receive an alarm duration prediction request about a predicted duration of a selected network alarm and alarm data regarding the selected network alarm and to generate an alarm duration prediction about a predicted duration of the selected network alarm by processing the alarm data regarding the selected network alarm through a first machine learning algorithm trained with alarm data regarding past network alarms gathered by the system, wherein:

the system generates the at least a subset of the gathered network alarms for generating the first set of network alarms based on the alarm duration predictions,

the selected network alarm is generated in response to faults/malfunctioning affecting a first network node of the telecommunication network, and

the alarm duration prediction is generated by processing, in addition to the alarm data regarding the selected network alarm, alarm data regarding prior network alarms generated in response to foregoing faults/malfunctioning affecting at least one among:

the first network node,

network nodes or links that are logically connected to the first network node, or

network nodes or links that are geographically located close to the first network node.

2. The system of claim 1 , wherein generating the alarm duration prediction comprises calculating, based on the alarm data regarding the selected network alarm, a duration probability indicative of a probability that the selected network alarm will have a duration shorter than a corresponding alarm duration threshold.

3. The system of claim 2 , wherein:

if the duration probability calculated is lower than a corresponding duration probability threshold, the system inserts the selected network alarm into the first set of network alarms; and

if the duration probability calculated is higher than or equal to the corresponding duration probability threshold, the system inserts the selected network alarm in a waiting queue for a time period corresponding to the alarm duration threshold, and then inserts the selected network alarm into the first set of network alarms if the selected network alarm has not subsided yet after expiration of the alarm duration threshold.

4. The system of claim 1 , wherein the system stores additional software that, when executed by the system, causes the system to:

receive a trouble ticket prediction request corresponding to the selected network alarm; and

generate a trouble ticket prediction by processing the alarm data regarding the selected network alarm to calculate a trouble ticket probability indicative of a probability of a need of issuing a Trouble ticket for solving faults/malfunctioning affecting a network node corresponding to the selected network alarm.

5. The system of claim 4 , wherein:

generating the selected network alarm comprises generating, in response to faults/malfunctioning affecting a first network node of the telecommunication network, the selected network alarm; and

generating the trouble ticket prediction comprises generating the trouble ticket prediction by processing, in addition to the alarm data regarding the selected network alarm, alarm data regarding prior network alarms generated in response to foregoing faults/malfunctioning affecting at least one among:

the first network node;

network nodes or links that are logically connected to the first network node; or

network nodes or links that are geographically located close to the first network node.

6. The system of claim 4 , wherein calculating the trouble ticket probability comprises processing the alarm data regarding the selected network alarm through a second machine learning algorithm trained with alarm data regarding past network alarms gathered by the system.

7. The system of claim 6 , wherein training the first machine learning algorithm comprises training the first machine learning algorithm independently from the second machine learning algorithm.

8. The system of claim 4 , where the system stores additional software that, when executed by the system, causes the system to:

if the trouble ticket probability is higher than or equal to a ticket probability threshold, generate a trouble ticket regarding the selected network alarm.

9. The system of claim 2 , the system stores additional software that, when executed by the system, causes the system to:

receive a trouble ticket prediction request corresponding to the selected network alarm and generate a trouble ticket prediction by processing the alarm data regarding the selected network alarm to calculate a trouble ticket probability indicative of a probability of a need of issuing a Trouble ticket for solving faults/malfunctioning affecting a network node corresponding to the selected network alarm, and

insert the selected network alarm in a waiting queue before adding the selected network alarm to the first set based on a combination of the duration probability and the trouble ticket probability.

10. The system of claim 1 , wherein the alarm data regarding a network alarm comprises at least one among:

data indicative of an active/inactive status of the network alarm;

data about a severity level of the network alarm among a set of predetermined severity levels;

data about geographical location of a network node having triggered the network alarm;

an identifier of the network node having triggered the network alarm;

data providing a category identifier of the network alarm; or

collected system logs regarding the network alarm.

11. A method for managing network alarms generated by element manager modules in response to corresponding faults/malfunctioning affecting network nodes of a telecommunication network, the method comprising:

gathering network alarms and generating a first set of network alarms comprising at least a subset of the gathered network alarms;

arranging the network alarms of the first set into a network alarm table providing, for each network alarm listed in the network alarm table, data describing the network alarm;

managing faults/malfunctioning affecting the network nodes corresponding to the network alarms based on the network alarm table; and

receiving an alarm duration prediction request about a predicted duration of a selected network alarm and alarm data regarding the selected network alarm and generating an alarm duration prediction about a predicted duration of the selected network alarm by processing the alarm data regarding the selected network alarm through a first machine learning algorithm trained with alarm data regarding past gathered network alarms, wherein:

the at least a subset of the gathered network alarms for generating the first set of network alarms is generated based on the alarm duration predictions,

the selected network alarm is generated in response to faults/malfunctioning affecting a first network node of the telecommunication network, and

the alarm duration prediction is generated by processing, in addition to the alarm data regarding the selected network alarm, alarm data regarding prior network alarms generated in response to foregoing faults/malfunctioning affecting at least one among:

the first network node,

network nodes or links that are logically connected to the first network node, or

network nodes or links that are geographically located close to the first network node.

12. The method of claim 11 , wherein generating the alarm duration prediction comprises calculating, based on the alarm data regarding the selected network alarm, a duration probability indicative of a probability that the selected network alarm will have a duration shorter than a corresponding alarm duration threshold.

13. The method of claim 12 , further comprising:

if the duration probability calculated is lower than a corresponding duration probability threshold, inserting the selected network alarm into the first set of network alarms; and

if the duration probability calculated is higher than or equal to the corresponding duration probability threshold, inserting the selected network alarm in a waiting queue for a time period corresponding to the alarm duration threshold, and then inserts the selected network alarm into the first set of network alarms if the selected network alarm has not subsided yet after expiration of the alarm duration threshold.

14. The method of claim 11 , further comprising:

receiving a trouble ticket prediction request corresponding to the selected network alarm; and

generating a trouble ticket prediction by processing the alarm data regarding the selected network alarm to calculate a trouble ticket probability indicative of a probability of a need of issuing a Trouble ticket for solving faults/malfunctioning affecting a network node corresponding to the selected network alarm.

15. The method of claim 14 , wherein:

generating the selected network alarm comprises generating, in response to faults/malfunctioning affecting a first network node of the telecommunication network, the selected network alarm; and

generating the trouble ticket prediction comprises generating the trouble ticket prediction by processing, in addition to the alarm data regarding the selected network alarm, alarm data regarding prior network alarms generated in response to foregoing faults/malfunctioning affecting at least one among:

the first network node;

network nodes or links that are logically connected to the first network node; or

network nodes or links that are geographically located close to the first network node.

16. The method of claim 14 , wherein calculating the trouble ticket probability comprises processing the alarm data regarding the selected network alarm through a second machine learning algorithm trained with alarm data regarding past network alarms.

17. The method of claim 16 , wherein training the first machine learning algorithm comprises training the first machine learning algorithm independently from the second machine learning algorithm.

18. The method of claim 14 , further comprising:

if the trouble ticket probability is higher than or equal to a ticket probability threshold, generating a trouble ticket regarding the selected network alarm.

19. The method of claim 12 , further comprising:

receiving a trouble ticket prediction request corresponding to the selected network alarm and generate a trouble ticket prediction by processing the alarm data regarding the selected network alarm to calculate a trouble ticket probability indicative of a probability of a need of issuing a Trouble ticket for solving faults/malfunctioning affecting a network node corresponding to the selected network alarm, and

inserting the selected network alarm in a waiting queue before adding the selected network alarm to the first set based on a combination of the duration probability and the trouble ticket probability.

20. The method of claim 11 , wherein the alarm data regarding a network alarm comprises at least one among:

data indicative of an active/inactive status of the network alarm;

data about a severity level of the network alarm among a set of predetermined severity levels;

data about geographical location of a network node having triggered the network alarm;

an identifier of the network node having triggered the network alarm;

data providing a category identifier of the network alarm; or

collected system logs regarding the network alarm.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (10)