Method and apparatus for classifying mobile network usage patterns
View Patent ↗In one embodiment, the method includes receiving, by a processor, subscriber usage data for a plurality of subscribers and performance metrics data for a plurality of cells of a wireless network. The method further includes identifying, by the processor, a set of subscribers that have one or more peak usage time intervals that are periodic. The method further includes identifying, by the processor, one or more peak usage time intervals of the plurality of cells in the wireless network. The method further includes classifying, by the processor, a first subset of the set of subscribers based on a correlation of the one or more peak usage time intervals of respective ones of the set of subscribers to the peak usage time intervals of the cell.
1. A method comprising:
receiving, by a processor, subscriber usage data for a plurality of subscribers and performance metrics data for a plurality of cells of a wireless network;
identifying, by the processor, a set of subscribers that have one or more peak usage time intervals that are periodic;
identifying, by the processor, one or more peak usage time intervals of the plurality of cells in the wireless network; and
classifying, by the processor, a first subset of the set of subscribers based on a correlation of the one or more peak usage time intervals of respective ones of the set of subscribers to the peak usage time intervals of the cell.
2. The method of claim 1 , wherein subscribers in the first subset are classified as high-value subscribers if the correlation for subscribers in the first subset is less than the correlation for subscribers of the set that are not in the first subset.
3. The method of claim 2 , further comprising:
offering service promotions to subscribers in the first subset.
4. The method of claim 1 , further comprising:
classifying a second subset of the set of subscribers as high-cost subscribers based on the correlation of the one or more peak usage time intervals of the respective ones of the set of subscribers to the peak usage time intervals of the cell.
5. The method of claim 4 , wherein the correlation for subscribers in the second subset is greater than the correlation of subscribers of the set that are not in the second subset.
6. The method of claim 5 , wherein subscribers in the second subset have peak usage time intervals that occur during peak usage time intervals of the cell.
7. The method of claim 5 , further comprising:
activating network resource control policies for subscribers in the second subset.
8. The method of claim 1 wherein the identifying the set of subscribers comprises:
determining periodicity of the peak usage time intervals of at least one of the plurality of subscribers by performing frequency domain analysis of usage data of at least one of the plurality of subscribers.
9. The method of claim 1 , wherein the classifying is performed for each cell of the plurality of cells.
10. An apparatus comprising:
a processor and associated memory configured to receive subscriber usage data for a plurality of subscribers and performance metrics data for a plurality of cells of a wireless network that serve the corresponding subscribers; and
the processor configured to identify a set of subscribers that have one or more peak usage time intervals that are periodic on the cells that serve the corresponding subscribers,
the processor configured to identify, for each cell of the plurality of cells, one or more peak usage time intervals of the cell in the wireless network, and
the processor configured to classify a first subset of the set of subscribers based on a correlation of the one or more peak usage time intervals of respective ones of the set of subscribers to the peak usage time intervals of the cell.
11. The apparatus of claim 10 , wherein subscribers in the first subset of the set of subscribers are classified as high-value subscribers if the correlation for subscribers in the first subset is less than the correlation for subscribers of the set that are not in the first subset.
12. The method of claim 11 , wherein the processor is further configured to offer service promotions to subscribers in the first subset.
13. The apparatus of claim 10 , wherein the processor is further configured to classify a second subset of the set of subscribers as high-cost subscribers based on the correlation of the one or more peak usage time intervals of the respective ones of the set of subscribers to the peak usage time intervals of the cell.
14. The apparatus of claim 13 , wherein the correlation for subscribers in the second subset of the set of subscribers is greater than the correlation of the set of subscribers that are not in the second subset.
15. The apparatus of claim 14 , wherein subscribers in the second subset of the set of subscribers have peak usage time intervals that occur during peak usage time intervals of the cell.
16. The apparatus of claim 10 , wherein the processor is further configured to activate network resource control policies for subscribers in the second subset.
17. The apparatus of claim 10 , wherein the processor determines periodicity of the peak usage time intervals of the plurality of subscribers by performing frequency domain analysis of the usage data of the plurality of subscribers.
18. The method of claim 10 , wherein the classifying is performed for each cell of the plurality of cells.