Anomaly correlation mechanism for analysis of handovers in a communication network

• US 9,918,258 B2
• Filed: 11/26/2013
• Issued: 03/13/2018
• Est. Priority Date: 11/26/2013
• Status: Active Grant

• Current Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Ehrlich, Willa Kay, Cui, Zhi, Phan, Duc, Wright, Marcus, Zhou, Qinggang
• Primary Examiner(s): Vu, Thai
• Application Number:
  US14/090,168
• Publication Number:
  US 20150148040A1
• Time in Patent Office: 1,568 Days
• CPC Class Codes: H04W 24/02 Arrangements for optimising operational condition
• CPC Class Codes: H04W 36/0079 in case of hand-off failure or rejection
• CPC Class Codes: H04W 36/00835 Determination of neighbour cell lists

Abstract

Optimizing neighbor cell relationships for improving handover performance by interpreting handover failures is presented herein. A method can include receiving data representing a time series of failures of outgoing handovers corresponding to a wireless access point device, and determining, based on a determined condition corresponding to a metric, that a source cell of the wireless access point device is associated with an anomalous cell relation corresponding to the time series of the failures of the outgoing handovers in response to quantifying the time series based on the metric. In various examples, the quantifying can include quantifying the time series with respect to: a volume of the failures during a period of time, an entropy calculated on a probability distribution of the failures by determined relations between the source cell and target cells, and/or an entropy calculated on a probability distribution of determined outgoing handover outcomes.

Claims

1. A method, comprising:

receiving, by a system comprising a processor, data representing a first time series of failures of outgoing handovers corresponding to a wireless access point device, wherein the outgoing handovers are outbound communication transfers from a source cell of the wireless access point device to target cells;

in response to quantifying the first time series of the failures of the outgoing handovers with respect to a first entropy that has been determined, based on a logarithm associated with the first time series of the failures, over a first probability distribution corresponding to the outgoing handovers, determining, by the system, that the source cell and a target cell of the target cells comprise a first anomalous relation represented by a first defined condition with respect to the first entropy; and

in response to the determining, sending, by the system, a message directed to a device identifying the first anomalous relation to facilitate removal of the first anomalous cell relation from a data store identifying the target cell as a handover candidate.

2. The method of claim 1, wherein the probability distribution represents a volume of the failures of the outgoing handovers during a period of time.

3. The method of claim 1, wherein the probability distribution represents the failures of the outgoing handovers by determined relations between the source cell and the target cells.

4. The method of claim 1, wherein the probability distribution represents determined outgoing handover outcomes.

5. The method of claim 1, wherein the determining comprises determining whether the first entropy, determined over the first probability distribution of the failures of the outgoing handovers by first determined relations between the source cell and the target cells, satisfies the first defined condition.

6. The method of claim 5, further comprising:

in response to determining that the first entropy satisfies the first defined condition, determining, by the system, that the first anomalous relation is a dominant relation using entropy based clustering.

7. The method of claim 5, further comprising:

in response to determining that the first entropy satisfies the first defined condition, determining, by the system, that the first anomalous relation is a dominant relation using non-parametric outlier detection.

8. The method of claim 5, further comprising:

in response to determining that the first anomalous relation is a dominant relation, receiving, by the system, incoming handover data representing a second time series of failures of incoming handovers corresponding to the target cell, wherein the incoming handovers are incoming communication transfers from source cells comprising the source cell to the target cell; and

determining, by the system, that the target cell and a selected source cell of the source cells comprise a second anomalous relation represented by a second defined condition with respect to the incoming handovers.

9. The method of claim 8, wherein the incoming handover data represents a volume of the failures of the incoming handovers during a period of time.

10. The method of claim 8, wherein the incoming handover data represents a second entropy that has been determined on a second probability distribution of the failures of the incoming handovers by determined relations between the target cell and the source cells.

11. The method of claim 8, wherein the incoming handover data represents a second entropy that has been determined on a second probability distribution of determined incoming handover outcomes.

12. The method of claim 8, further comprising:

in response to the determining the second anomalous relation, correlating, by the system, the first anomalous relation with the second anomalous relation.

13. A system, comprising:

a processor; and

a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations, comprising;

receiving data representing a time series of failures of outgoing handovers corresponding to a source device, wherein the outgoing handovers comprise outbound communication transfers from a source cell of the source device to target cells;

in response to quantifying the time series based on a standardized entropy that has been determined over a probability distribution of the outgoing handovers, determining that an anomalous relation comprising the source cell and a target cell of the target cells satisfies a defined condition with respect to the standardized entropy; and

removing the anomalous relation from a data store identifying the target cell as an outgoing handover candidate.

14. The system of claim 13, wherein the probability distribution represents a volume of the failures of the outgoing handovers during a period of time.

15. The system of claim 13, wherein the probability distribution represents the failures of the outgoing handovers by determined relations between the source cell and the target cells.

16. The system of claim 13, wherein the probability distribution represents determined outgoing handover outcomes.

17. The system of claim 13, wherein the defined condition represents the failures of the outgoing handovers by determined relations between the source cell and the target cells.

18. A non-transitory machine-readable storage medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, comprising:

receiving data representing a time series of failures of outgoing handovers corresponding to a source device, wherein the outgoing handovers comprise outbound communication transfers from a source cell of the source device to target cells;

in response to quantifying the time series based on an entropy that has been derived, based on a logarithm corresponding to the time series of failures, over a probability distribution of the outgoing handovers, identifying that the source cell corresponds to cells associated with an anomalous handover failure condition, wherein the cells comprise the source cell and a target cell of the target cells; and

removing the source cell and the target cell from a data store identifying the target cell as a candidate for an outgoing handover.

19. The non-transitory machine-readable storage medium of claim 18, wherein the probability distribution represents the failures of the outgoing handovers by determined relations between the source cell and the target cells.

20. The non-transitory machine-readable storage medium of claim 19, wherein the identifying comprises:

in response to determining that the entropy satisfies a determined metric with respect to determined relations between the source cell and the target cells, identifying that the cells is an anomalous relation using entropy based clustering.