Resource budget determination for communications network

Patent metadata

Field	Value
Publication number	US-10135698-B2
Application number	US-201314890968-A
Country	US
Kind code	B2
Filing date	May 14, 2013
Priority date	May 14, 2013
Publication date	Nov 20, 2018
Grant date	Nov 20, 2018

Abstract

Resources (10, 20, 30) of a communications network (40), are used to provide services, some are superordinate services (A, B) which depend on contributory services (C, D, E, F). Key performance indicators R-KPI are received based on measurements from the resources, of contributory services contributing to a corresponding key performance indicator KPI of the corresponding superordinate service. A selection (120) is made of the received R-KPIs which correspond to the normal operation of the superordinate service and a characteristic division is generated (130) of relative contributions of the R-KPIs to the KPI of the corresponding superordinate service for the times that the service was operating normally. This can be used as a basis for identifying differences with other R-KPIs to determine anomalies in behavior of contributory services for root cause analysis or in managing the network. Convergence evaluation and validation can be used to improve accuracy of the characteristic division.

First claim

The invention claimed is:

1. A method of monitoring a communications network having resources used to provide services, some of which are superordinate services which depend on contributory ones of the services, the method comprising:
selecting from a plurality of resource key performance indicators, R-KPIs, obtained from measurements from the resources, the R-KPIs being of contributory services contributing to a key performance indicator, KPI, of a corresponding superordinate service, wherein selecting is based on indications of whether the superordinate service was operating normally at times related to received R-KPIs, so as to select those of the received R-KPIs which correspond to a normal operation of the superordinate service;
generating automatically from the selected R-KPIs a characteristic division of relative contributions of the R-KPIs to the KPI of the corresponding superordinate service for the times that the service was operating normally; and
validating the generated characteristic division by identifying a mismatch between a first state of whether current R-KPIs are individually acceptable and a second state of whether a current KPI of the corresponding superordinate service is acceptable.

2. The method of claim 1 , further comprising identifying differences between at least R-KPIs other than those selected, and the characteristic division, and determining behavior of the contributory services relating to these R-KPIs based on the identified differences.

3. The method of claim 2 , further comprising managing the network by managing at least one of the contributory services, based on the identified differences.

4. The method of claim 2 , for use in a root cause analysis of a change in behavior of the superordinate service, and wherein identifying the differences is carried out for R-KPIs related to the change in behavior of the superordinate service and comprises determining the root cause based on the determined behavior of the contributory services related to these R-KPIs.

5. The method of claim 4 , further comprising identifying which of the R-KPIs are associated with a time of the change, and selecting these associated R-KPIs for identifying the differences.

6. The method of claim 1 , wherein selecting those of the received R-KPIs which correspond to the normal operation of the superordinate service comprises determining a correspondence between at least one of the received R-KPIs and the times of normal operation based on an expected delay in how the respective contributory service affects the superordinate service.

7. The method of claim 1 , further comprising determining whether values of the KPI of the corresponding superordinate service, obtained from measurements of the superordinate service, are acceptable, to indicate whether the superordinate service was operating normally.

8. The method of claim 1 , further comprising:
evaluating convergence of the characteristic division, by generating a sequence of candidate characteristic divisions for different time periods, and
evaluating how much the sequence of candidate characteristic divisions converges towards an optimal candidate division.

9. The method of claim 1 , wherein the R-KPIs comprise at least one of: delays, latency, loss ratio, noise, bit errors, packet errors, packet losses, frame errors, frame losses, amount of variation of any of these.

10. An apparatus for monitoring a communications network, the communications network comprising resources for use in providing services, some of the services being superordinate services which depend on contributory ones of the services, the apparatus comprising:
a selector configured to select from a plurality of resource key performance indicators, R-KPIs, obtained from measurements from the resources, of contributory services contributing to a key performance indicator, KPI, of a corresponding superordinate service, wherein selecting is based on an indication of whether the superordinate service was operating normally at the time, so as to select R-KPIs which correspond to a normal operation of the superordinate service, and
a division generator configured to generate from the selected R-KPIs a characteristic division of relative contributions of the R-KPIs to the KPI of the corresponding superordinate service for the times that the superordinate service was operating normally;
wherein the apparatus is configured to validate the generated characteristic division by identifying a mismatch between a first state of whether current R-KPIs are individually acceptable and a second state of whether a current KPI of the corresponding superordinate service is acceptable.

11. The apparatus of claim 10 , further comprising a difference analyser configured to identify differences between R-KPIs other than those selected, and the characteristic division, and configured to determine behavior of the contributory services based on the identified differences.

12. The apparatus of claim 11 , configured for root cause analysis of a change in behavior of the superordinate service, by having the selector configured to select R-KPIs related to the change in behavior of the superordinate service for input to the difference analyser, and further comprising a root cause analyser configured to receive an indication of the determined behavior of the contributory services related to these R-KPIs, and to determine a root cause based on this indication.

13. The apparatus of claim 12 wherein the selector for selecting R-KPIs related to the change in behavior is also configured to identify which of the R-KPIs are associated with a time of the change, and to select these associated R-KPIs for use in identifying the differences.

14. The apparatus of claim 11 , further comprising service controller configured to manage at least one of the contributory services, based on the identified differences.

15. The apparatus according to claim 10 , further comprising a processor and a memory, wherein at least one of the selector, division generator, difference analyzer, root cause analyzer or service controller is implemented in software operating in the processor based on instructions stored in the memory.

16. A communication network comprising resources for use by services, and having an apparatus monitoring a communications network, the communications network comprising resources for use in providing services, some of the services being superordinate services which depend on contributory ones of the services, the apparatus comprising:
a selector configured to select from a plurality of resource key performance indicators, R-KPIs, obtained from measurements from the resources, of contributory services contributing to a key performance indicator, KPI, of a corresponding superordinate service, wherein selecting is based on an indication of whether the superordinate service was operating normally at the time, so as to select R-KPIs which correspond to a normal operation of the superordinate service, and
a division generator configured to generate from the selected R-KPIs a characteristic division of relative contributions of the R-KPIs to the KPI of the corresponding superordinate service for the times that the superordinate service was operating normally;
wherein the apparatus is configured to validate the generated characteristic division by identifying a mismatch between a first state of whether current R-KPIs are individually acceptable and a second state of whether a current KPI of the corresponding superordinate service is acceptable.

Assignees

• Ericsson Telefon Ab L M
• Ericsson Telefon Ab L M

Inventors

• Groenendijk Jan
• Granath Johan Christer Wilhelm
• Huang Yangcheng

Classifications

• H04L41/5009 Primary

  Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF] · CPC title

• H04L41/0631

  using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis · CPC title

• H04L41/5038

  Electricity · mapped topic

• H04L41/5035

  Electricity · mapped topic

• H04L43/55

  Testing of service level quality, e.g. simulating service usage · CPC title