Detecting communication network insights of alerts

Patent metadata

Field	Value
Publication number	US-11258709-B2
Application number	US-202016930712-A
Country	US
Kind code	B2
Filing date	Jul 16, 2020
Priority date	Jul 17, 2018
Publication date	Feb 22, 2022
Grant date	Feb 22, 2022

Abstract

In one embodiment, the system identifies one or more geographic areas covered by a communication network. The system determines, for each identified geographic area, a congestion metric for the identified geographic area based at least on a difference between a first and second reference point on a network speed curve, wherein the network speed curve represents download speeds for a volume of traffic in the identified geographic area. The system identifies one or more network traffic congestions in one or more of the identified geographic areas based on a comparison of the respective congestion metrics of the identified geographic areas to a threshold congestion metric. The system sends, to one or more operators of the communication network, one or more alerts about the identified network traffic congestions.

First claim

What is claimed is:

1. A method comprising, by a computing system:
identifying one or more geographic areas covered by a communication network;
determining, for each geographic area, a congestion metric for the geographic area based at least on a difference between a first and second reference point on a network speed curve, wherein the network speed curve represents download speeds for a volume of traffic in the geographic area;
identifying one or more network traffic congestions in one or more of the geographic areas based on a comparison of the respective congestion metrics of the geographic areas to a threshold congestion metric; and
sending, to one or more operators of the communication network, one or more alerts about the identified network traffic congestions.

2. The method of claim 1 , wherein the congestion metric is based on a ratio of the difference between the first and second reference points to the first reference point.

3. The method of claim 1 , wherein the first reference point is based on the download speed of a busy-hour volume of traffic, and wherein the second reference point is based on the download speed of a non-busy-hour volume of traffic.

4. The method of claim 3 , wherein the download speed of the busy-hour volume of traffic is a mean, median, or mode of the download speed of the busy hours, and wherein the download speed of the non-busy-hour volume of traffic is a mean, median, or mode of the of the download speed of the non-busy hours.

5. The method of claim 1 , further comprising:
ranking the one or more geographic areas based on a congestion severity indicated by the respective congestion metrics of the one or more geographic areas; and
sending the one or more alerts with the one or more geographic areas ranked based on the congestion metrics.

6. The method of claim 1 , further comprising:
filtering the one or more alerts; and
prioritizing, based on the filtered one or more alerts, an optimization of a network performance to increase a capacity of the communication network in the one or more geographic areas.

7. The method of claim 6 , wherein the one or more alerts are filtered by one or more of:
usage cases;
congested one or more geographic areas;
one or more geographic areas with coverage issues;
problem devices; or
cost to fix.

8. The method of claim 6 , wherein the one or more alerts are filtered based on an occurrence rate of the one or more alerts, and wherein the occurrence rate of the one or more alerts can be determined by taking a ratio of the one or more alerts over a period of time.

9. The method of claim 1 , further comprising:
updating an alert database with the one or more alerts about the identified network traffic congestions.

10. The method of claim 1 , further comprising:
performing a root cause analysis for the identified network traffic congestions.

11. The method of claim 10 , wherein the root cause analysis is performed using a machine learning model that is trained based on historical data of the identified network traffic congestions.

12. The method of claim 1 , wherein the threshold congestion metric is determined by a congestion-analysis machine learning model.

13. The method of claim 12 , wherein the threshold congestion metric is determined by a classification model or a tree model associated with the congestion-analysis machine learning model trained by manually labeled data, and wherein the threshold congestion metric is adjusted by balancing a precision metric and a recall rate of the identified one or more network traffic congestions.

14. The method of claim 1 , wherein the first and second reference points on the network speed curve represent average download speeds for the volume of traffic.

15. The method of claim 14 , wherein the average download speed is a first mean download speed in one or more first prior time periods based on a plurality of aggregated download speeds in one or more first prior time periods, and wherein each aggregated speed in the one or more first prior time periods is a second mean download speed in a plurality of days.

16. The method of claim 14 , wherein the first average download speed is a first median download speed in one or more first prior time periods based on a plurality of aggregated download speeds in one or more first prior time periods, and wherein each aggregated network speed in the one or more first prior time periods is a second median download speed in a plurality of days.

17. The method of claim 14 , wherein the first average download speed is a mean download speed in the one or more first time periods, and wherein the second average download speed is a mean download speed in one or more second time periods.

18. The method of claim 14 , wherein the first average download speed is a median download speed in the one or more first prior time periods, and wherein the second average download speed is a median download speed in one or more second prior time periods.

19. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
identify one or more geographic areas covered by a communication network;
determine, for each geographic area, a congestion metric for the geographic area based at least on a difference between a first and second reference point on a network speed curve, wherein the network speed curve represents download speeds for a volume of traffic in the geographic area;
identify one or more network traffic congestions in one or more of the geographic areas based on a comparison of the respective congestion metrics of the geographic areas to a threshold congestion metric; and
send, to one or more operators of the communication network, one or more alerts about the identified network traffic congestions.

20. A system comprising: one or more non-transitory computer-readable storage media embodying instructions; and one or more processors coupled to the storage media and operable to execute the instructions to:
identify one or more geographic areas covered by a communication network;
determine, for each geographic area, a congestion metric for the geographic area based at least on a difference between a first and second reference point on a network speed curve, wherein the network speed curve represents download speeds for a volume of traffic in the geographic area;
identify one or more network traffic congestions in one or more of the geographic areas based on a comparison of the respective congestion metrics of the geographic areas to a threshold congestion metric; and
send, to one or more operators of the communication network, one or more alerts about the identified network traffic congestions.

Assignees

• Meta Platforms Inc

Inventors

• Li Ying
• De Jongh Martinus Arnold
• Gonguet Vincent
• Donovan James
• Corigliano Laurent

Classifications

• H04L47/12

  Avoiding congestion; Recovering from congestion · CPC title

• H04L67/535

  Tracking the activity of the user (network monitoring arrangements H04L43/00; recording of computer activity G06F11/34) · CPC title

• H04L41/5009

  Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF] · CPC title

• H04L43/0894

  Packet rate · CPC title

• H04L47/11 Primary

  Identifying congestion · CPC title