Geospatial telecommunications network coverage balancing using network cell boundaries
Data describing network coverage provided to multiple user equipments by a telecommunications network is obtained. The data is associated with geographical locations serviced by the telecommunications network. The data is grouped according to the geographical locations into multiple clusters. Using the multiple clusters, multiple geospatial polygons associated with cells of the telecommunications network are generated. Differences are determined in the network coverage between the cells using the geospatial polygons. Imbalances are determined in the telecommunications network based on the differences in the network coverage between the cells.
1. A computer system comprising:
at least one hardware processor; and
at least one non-transitory memory storing instructions, which, when executed by the at least one hardware processor, cause the computer system to:
obtain network data describing network coverage provided to multiple user equipments by a sector of a cell site including a first cell and a second cell,
wherein the network data is associated with multiple geographical locations serviced by the sector, and
wherein the first cell and the second cell are associated with a particular spectrum frequency of the sector;
group, using an unsupervised artificial intelligence method, the network data according to the geographical locations into a first cluster and a second cluster,
wherein the first cluster is associated with the first cell and the second cluster is associated with the second cell;
generate a first geospatial polygon associated with the first cluster and a second geospatial polygon associated with the second cluster;
determine an imbalance in the network coverage by comparing the first geospatial polygon with the second geospatial polygon; and
perform an action on the cell site to adjust the imbalance in the network coverage.
2. The computer system of claim 1 , wherein the computer system is caused to:
train an artificial intelligence model using the network data to:
determine the imbalance in the network coverage; and
determine the action to adjusting the imbalance in the network coverage.
3. The computer system of claim 1 , wherein the network data is obtained by crowdsourcing.
4. The computer system of claim 1 , wherein performing the action includes adjusting a tilt of an antenna of the cell site.
5. The computer system of claim 1 , wherein performing the action includes adjusting an azimuth of an antenna of the cell site.
6. The computer system of claim 1 , wherein the first cell has a first bandwidth at the particular spectrum frequency and the second cell has a second bandwidth at the particular spectrum frequency, and
wherein the computer system is caused to:
compare traffic carried by the first cell and the second cell based on the first bandwidth and the second bandwidth.
7. The computer system of claim 1 , wherein the first geospatial polygon is a convex hull, and
wherein generating the convex hull includes connecting endpoints of the first cluster.
8. At least one non-transitory computer-readable storage medium storing instructions, which, when executed by at least one data processor of a computer system, cause the computer system to:
obtain network data associated with a sector of a cell site;
group the network data into a first cluster associated with a first layer and a second cluster associated with a second layer,
wherein the first layer is associated with a first spectrum frequency, and the second layer is associated with a second spectrum frequency greater than the first spectrum frequency;
determine, from the network data, a first amount of traffic associated with the first cluster and a second amount of traffic associated with the second cluster;
determine, using an artificial intelligence model, an imbalance in the traffic based on the first amount of traffic and the second amount of traffic; and
determine an action to adjust the imbalance in the traffic.
9. The non-transitory computer-readable storage medium of claim 8 , wherein the first layer is associated with a first bandwidth and the second layer is associated with a second bandwidth greater than the first bandwidth.
10. The non-transitory computer-readable storage medium of claim 8 , wherein the second amount of traffic is greater than the first amount of traffic.
11. The non-transitory computer-readable storage medium of claim 8 , wherein the first layer is associated with a first network coverage and the second layer is associated with a second network coverage less than the first network coverage.
12. The non-transitory computer-readable storage medium of claim 8 , wherein the computer system is caused to:
generate a first geospatial polygon associated with the first cluster and a second geospatial polygon associated with the second cluster; and
determine whether the second geospatial polygon lies within the first geospatial polygon.
13. The non-transitory computer-readable storage medium of claim 12 , wherein the first geospatial polygon circumscribes a first cluster boundary, and the second geospatial polygon circumscribes a second cluster boundary.
14. The non-transitory computer-readable storage medium of claim 8 , wherein grouping the network data removes outliers from the network data.
15. A method performed by a computer system, the method comprising:
obtaining network data describing network coverage provided by a sector of a cell site;
grouping the network data into multiple clusters associated with multiple layers of the sector, wherein each layer of the multiple layers has multiple frequency carriers;
generating multiple convex hulls based on the clusters;
determining differences in the network coverage across the frequency carriers using the convex hulls;
determining an imbalance in the sector based on the differences in the network coverage; and
determining an action to adjust the imbalance in the sector based on the differences in the network coverage.
16. The method of claim 15 , wherein determining the differences in the network coverage includes comparing boundaries of the convex hulls.
17. The method of claim 15 , wherein the layers include at least two of a low-band 5G layer, a mid-band 5G layer, and a high-band-5G layer.
18. The method of claim 15 , comprising:
performing the action on the cell site; and
determining whether performing the action increased a key performance indicator associated with the cell site.
19. The method of claim 15 , comprising:
training an artificial intelligence model, using the network data to:
determine the imbalance in the sector.
20. The method of claim 15 , wherein determining the differences in the network coverage is accomplished without drive testing, preventing greenhouse gas emissions caused by multiple vehicles used to determine the imbalance in the sector.