Network fault detection and quality of service improvement systems and methods
A Quality of Service (QoS) improvement system and method for transport network fault detection and QoS improvement so that a telecommunication network service provider can analyze the root cause on chronic performance issues and recommend potential solutions is disclosed. The system runs performance analysis on each AAV (mobile backhaul) or other transport networks and collects performance related metrics data. The system then selects a subset of the data related to certain key performance indicators (KPIs), such as latency, jitter, packet loss ratio, and availability. On this subset of KPI-related data, the system applies clustering techniques to identify clusters with similar performance issues. For each cluster, the system binds the AAV performance KPI data with one or more of the following site features—health, location, vendor, market, service type, etc.—to create a cluster map. The system can then generate inferences on root causes of the performance issues.
1. A computer-implemented method for detecting faults in a carrier network, the method comprising:
receiving a set of performance measurement data records measuring performance of at least one network access component;
generating a set of clusters of the set of performance measurement data records using values corresponding to one or more key performance indicators measuring the performance of the at least one network access component; and
updating each cluster in the set of clusters with wireless site information to enable determination of similarity of service areas associated with said each cluster of the set of clusters.
2. The method of claim 1 , wherein the set of clusters is generated using k-means clustering, and wherein an optimal k value is determined using an elbow methodology.
3. The method of claim 1 , wherein the set of clusters is generated using k-means clustering, and wherein an optimal k value is determined using a silhouette methodology.
4. The method of claim 1 , further comprising:
receiving the set of performance measurement data records for every time period t;
generating mapped cluster records for one or more of the set of performance measurement data records received for said each time period t; and
analyzing the mapped cluster records to identify at least one network fault trend in the carrier network.
5. The method of claim 1 , further comprising:
mapping the updated clusters with at least one clustering parameter to generate mapped updated clusters of the data records; and
identifying at least one network fault occurring in the carrier network using the generated mapped updated cluster data records.
6. The method of claim 1 , further comprising:
mapping the updated clusters with at least one clustering parameter to generate mapped updated clusters of the data records; and
identifying at least one network fault remediation action capable of being performed in the carrier network based on the generated mapped updated cluster data records.
7. The method of claim 1 , wherein the key performance indicators comprises one or more of:
latency,
jitter,
packet loss ratio, or
availability.
8. At least one non-transitory computer-readable medium containing instructions, that when executed by a processor, performs operations for detecting faults in a carrier network, the operations comprising:
receiving a set of performance measurement data records measuring performance of at least one network access component;
generating a set of clusters of the set of performance measurement data records using values corresponding to one or more key performance indicators measuring the performance of the at least one network access component; and
updating each cluster in the set of clusters with wireless site information to enable determination of similarity of service areas associated with said each cluster of the set of clusters.
9. The at least one non-transitory computer-readable medium of claim 8 , wherein the set of clusters is generated using k-means clustering, and wherein an optimal k value is determined using an elbow methodology.
10. The at least one non-transitory computer-readable medium of claim 8 , wherein the set of clusters is generated using k-means clustering, and wherein an optimal k value is determined using a silhouette methodology.
11. The at least one non-transitory computer-readable medium of claim 8 , wherein the operations further comprise:
receiving the set of performance measurement data records for every time period t;
generating mapped cluster records for one or more of the set of performance measurement data records received for said each time period t; and
analyzing the mapped cluster records to identify at least one network fault trend in the carrier network.
12. The at least one non-transitory computer-readable medium of claim 8 , wherein the operations further comprise:
mapping the updated clusters with at least one clustering parameter to generate mapped updated clusters of the data records; and
identifying at least one network fault occurring in the carrier network using the generated mapped updated cluster data records.
13. The at least one non-transitory computer-readable medium of claim 8 , wherein the operations further comprise:
mapping the updated clusters with at least one clustering parameter to generate mapped updated clusters of the data records; and
identifying at least one network fault remediation action capable of being performed in the carrier network based on the generated mapped updated cluster data records.
14. The at least one non-transitory computer-readable medium of claim 8 , wherein the key performance indicators comprises one or more of:
latency,
jitter,
packet loss ratio, or
availability.
15. A system for detecting faults in a carrier network comprising at least one non-transitory computer-readable medium having instructions stored thereon, which when executed by one or more processors of the system cause the system to:
receiving a set of performance measurement data records measuring performance of at least one network access component;
generating a set of clusters of the set of performance measurement data records using values corresponding to one or more key performance indicators measuring the performance of the at least one network access component; and
updating each cluster in the set of clusters with wireless site information to enable determination of similarity of service areas associated with said each cluster of the set of clusters.
16. The system of claim 15 , wherein the set of clusters is generated using k-means clustering, and wherein an optimal k value is determined using an elbow methodology or a silhouette methodology.
17. The system of claim 15 , wherein the instructions when executed by the one or more processors of the system further cause the system to:
receive the set of performance measurement data records for every time period t;
generate mapped cluster records for one or more of the set of performance measurement data records received for said each time period t; and
analyze the mapped cluster records to identify at least one network fault trend in the carrier network.
18. The system of claim 15 , wherein the instructions when executed by the one or more processors of the system further cause the system to:
map the updated clusters with at least one clustering parameter to generate mapped updated clusters of the data records; and
identify at least one network fault occurring in the carrier network using the generated mapped updated cluster data records.
19. The system of claim 15 , wherein the instructions when executed by the one or more processors of the system further cause the system to:
map the updated clusters with at least one clustering parameter to generate mapped updated clusters of the data records; and
identify at least one network fault remediation action capable of being performed in the carrier network based on the generated mapped updated cluster data records.
20. The system of claim 15 , wherein the key performance indicators comprises one or more of:
latency,
jitter,
packet loss ratio, or
availability.