Method and system for isolating related events in the presence of seasonal variations
This disclosure provides a method and system that isolate non-seasonal components of network measurements from the seasonal components and compare the non-seasonal components to determine cause and effect relationships between various network nodes, to determine whether an event at a first node is experienced at a second node, and/or to characterize and localize a source of interference.
1. A method for a wireless telecommunications network, the method comprising:
receiving first event measurements of a first event that were measured at a first node of a wireless communication network over a time interval, wherein the first event is caused by non-cellular interference;
receiving second event measurements of a second event that were measured at a second node of the wireless communication network over the time interval;
removing seasonal components from the first event measurements to create first non-seasonal event data;
removing seasonal components from the second event measurements to create second non-seasonal event data;
comparing the first non-seasonal event data with the second non-seasonal event data to determine whether the first event measured by the first node correlates with the second event measured at the second node;
determining which cells are affected by the non-cellular interference; and
using data for the affected cells to locate a source of the non-cellular interference.
2. The method of claim 1 , wherein, when the first event experienced by the first node correlates with the second event measured at the second node, determining that the first node and the second node are both affected by the same interference.
3. The method of claim 2 , wherein the first and second nodes are co-sited cells, the method further comprising:
receiving interference measurements for a plurality of cells that are neighbors of one or both of the co-sited cells;
removing seasonal components from the neighbor interference measurements to create non-seasonal neighbor interference data; and
comparing at least one of the first non-seasonal event data and the second non-seasonal event data with the non-seasonal neighbor interference data to determine whether a correlation is present.
4. The method of claim 3 , further comprising:
when no correlation is present, determining that interference experienced by the co-sited cells is local to a cell site of the co-sited cells.
5. The method of claim 1 , wherein the first non-seasonal event data is a trend component of the first event data, and the second non-seasonal event data is a trend component of the second event data.
6. The method of claim 1 , further comprising:
measuring the first event at the first node; and
measuring the second event at the second node.
7. The method of claim 1 , further comprising:
performing time-series decomposition on the first and second measurement data to remove the seasonal components.
8. A method for a wireless telecommunications network, the method comprising:
measuring a first event that were measured at a first node of a wireless communication network over a time interval, wherein the first event is caused by non-cellular interference;
measuring a second event at a second node of the wireless communication network over the time interval;
removing seasonal components from the first event measurements to create first non-seasonal event data;
removing seasonal components from the second event measurements to create second non-seasonal event data;
comparing the first non-seasonal event data with the second non-seasonal event data to determine whether the first event measured by the first node correlates with the second event measured at the second node;
determining which cells are affected by the non-cellular interference; and
using data for the affected cells to locate a source of the non-cellular interference.
9. The method of claim 8 , wherein, when the first event experienced by the first node correlates with the second event measured at the second node, determining that the first node and the second node are both affected by the same interference.
10. The method of claim 8 , wherein the first and second nodes are co-sited cells, the method further comprising:
receiving interference measurements for a plurality of cells that are neighbors of one or both of the co-sited cells;
removing seasonal components from the neighbor interference measurements to create non-seasonal neighbor interference data; and
comparing at least one of the first non-seasonal event data and the second non-seasonal event data with the non-seasonal neighbor interference data to determine whether a correlation is present.
11. The method of claim 10 , further comprising:
when no correlation is present, determining that interference experienced by the co-sited cells is local to a cell site of the co-sited cells.
12. The method of claim 8 , wherein the first non-seasonal event data is a trend component of the first event data, and the second non-seasonal event data is a trend component of the second event data.
13. A non-transitory computer-readable medium with computer-executable instructions stored thereon which, when executed by a processor, perform the following operations:
accessing first event measurements of a first event that were measured at a first node of a wireless communication network over a time interval, wherein the first event is caused by non-cellular interference;
accessing second event measurements of a second event that were measured at a second node of the wireless communication network over the time interval;
removing seasonal components from the first event measurements to create first non-seasonal event data;
removing seasonal components from the second event measurements to create second non-seasonal event data;
comparing the first non-seasonal event data with the second non-seasonal event data to determine whether the first event measured by the first node correlates with the second event measured at the second node;
determining which cells are affected by the non-cellular interference; and
using data for the affected cells to locate a source of the non-cellular interference.