Systems and methods for network monitoring and testing using self-adaptive triggers based on KPI values
Embodiments are directed to systems and methods for network monitoring and testing using self-adaptive Key Performance Indicator to establish thresholds to trigger alarms in network monitoring. The present embodiments will be described in the context of calculating KPIs for call detail records (CDRs) and using the KPIs for network monitoring.
1. A method of determining a normal key performance indicator (KPI) value, comprising the steps of:
recording an initial set of a plurality of KPI values during a learning period;
computing a normal KPI value based on the initial set of the plurality of KPI values at an end of the learning period;
monitoring the plurality of KPI values;
comparing one KPI value of the plurality of KPI values to a range of normal KPI values, wherein the one KPI value is indicative of a network performance issue;
sending an alarm to one or more recipients based on determination of a deviation between the one KPI value to the range of KPI values and based on a first set of rules describing conditions for sending the alarm to the one or more recipients based on the deviation; and
using one or more KPI values from the initial set of the plurality of KPI values and one or more new KPI values from the plurality of KPI values to update the normal KPI value based on a second set of rules describing conditions for:
updating the normal KPI value; and
determining which of the one or more new KPI values should be used to update the normal KPI value and which of the one or more new KPI values should be excluded from updating the normal KPI;
wherein:
sending the alarm to the one or more recipients comprises sending the alarm over a packet-based network; and
repeating said monitoring, comparing, and sending using the updated normal KPI value.
2. The method of claim 1 , wherein the normal KPI value, the initial set of the plurality of KPI values, and the plurality of KPI values are associated with events of a network.
3. The method of claim 1 , wherein the normal KPI value, the initial set of the plurality of KPI values, and the plurality of KPI values are indicative of metrics associated with quality of experience over a network.
4. The method of claim 1 , wherein the normal KPI value is associated with a time frame, and wherein the one KPI value is compared to the range of normal KPI values based on the time frame.
5. The method of claim 1 , wherein the comparison of the one KPI value to the range of normal KPI values comprises determining a deviation between the one KPI value and the range of normal KPI values and a direction of the deviation.
6. The method of claim 5 , wherein the alarm is sent based on the direction of the deviation.
7. The method of claim 5 , wherein a user-defined rule specifies whether the alarm is to be sent based on the direction of the deviation.
8. The method of claim 1 , further comprises presenting a list of KPIs for monitoring to a user and receiving from the user a selection of KPIs to be monitored.
9. The method of claim 8 , wherein for each of the KPIs from the selection of KPIs, a corresponding normal KPI value is automatically computed and is used in generating an alarm.
10. The method of claim 1 , further comprises providing an interface for displaying the normal KPI value.
11. The method of claim 1 , wherein the first set of rules comprise default deviation values for KPIs for triggering alarms.
12. The method of claim 1 , wherein the alarm triggers a sequence of steps for mitigating an issue associated with the one KPI value.
13. The method of claim 12 , wherein the one KPI value is indicative of a performance of a network device, and wherein the sequence of steps reroutes a network traffic to avoid the network device.
14. The method of claim 1 , wherein the first set of rules define content of the alarm and a type of delivery of the alarm.
15. The method of claim 1 , wherein the second set of rules allow an update of the normal KPI value when the one or more KPI values from the initial KPI values and the one or more new KPI values are within a range of the normal KPI value.
16. The method of claim 1 , wherein the first set of rules and the second set of rules allow the alarm to be triggered based on the comparison of the one KPI value to the range of normal KPI values and prohibit the one KPI value from being used in updating the normal KPI value.
17. The method of claim 1 , wherein the learning period accounts for periodic network events and permanent network events associated with the initial set of the plurality of KPI values and the one or more new KPI values.
18. The method of claim 1 , wherein the learning period is defined by a user.
19. The method of claim 1 , wherein the learning period is automatically set based on a predicted network event and wherein the first rule is automatically updated based on the predicted network event.
20. The method of claim 1 , wherein the second set of rules further comprise a weighting algorithm for updating the normal KPI value.
21. The method of claim 1 , wherein the computing a normal KPI value comprises using an average and a deviation of the initial set of the plurality of KPI values.
22. The method of claim 1 , wherein the computing a normal KPI value comprises computing a range associated with the normal KPI value, wherein the range is based on a maximum and minimum from the initial set of the plurality of KPI values.
23. The method of claim 1 , wherein the recording an initial plurality of KPI values comprises transforming the initial set of the plurality of KPI values by applying a transform function to the plurality of KPI values, and wherein the normal KPI value is expressed using the transform function.
24. A method of determining a normal key performance indicator (KPI) value, comprising the steps of:
recording an initial set of a plurality of KPI values during a learning period;
computing a normal KPI value based on the initial set of the plurality of KPI values at an end of the learning period;
monitoring the plurality of KPI values;
comparing one KPI value of the plurality of KPI values to a range of normal KPI values, wherein the one KPI value is indicative of a network performance issue;
sending an alarm to one or more recipients based on a comparison of the one KPI value to the range of KPI values and based on a first set of rules describing conditions for sending the alarm to the one or more recipients based on the comparison, wherein the alarm is usable to trigger a sequence of steps for mitigating the network performance issue; and
using one or more KPI values from the initial set of the plurality of KPI values and one or more new KPI values from the plurality of KPI values to update the normal KPI value based on a second set of rules describing conditions for:
updating the normal KPI value; and
determining which of the one or more new KPI values should be used to update the normal KPI value and which of the one or more new KPI values should be excluded from updating the normal KPI;
wherein:
sending the alarm to the one or more recipients comprises sending the alarm over a packet-based network; and
repeating said monitoring, comparing, and sending using the updated normal KPI value.