Telecom
Industry
Objective, Scope, and Collected Data
Objective
The objective of the PoC is to identify an advanced analytic solution (PFM) that can help answer these questions:
- Which ones have a higher probability to generate faults in the upcoming period?
- Which are the critical alarms that have a greater probability to occur in the upcoming period?
- What are the correlations between traffic levels, characteristics of the equipment, fault history, and upcoming occurrence of a fault/alarm?
Scope
A subset of the Core 3G Network (limited to MSS/MGW-OMGW, SGW, and HLR).
Collected Data
- Operational/Traffic data (volumes, calls, users, restarts, etc.)
- Thresholds alarms (alarms with critical severity)
- Equipment / device / components (device name, location and attributes)
- Performance logs (available only for MSS)
- Swap (part replacement) occurred on physical devices.
(These are the faults in scope to the analysis)
Business Benefits
Faults
In terms of business, value anticipated faults are the ones that will allow the customer to more effectively manage their network
Short term predictions:
- Adjust Network operations weekly maintenance priorities.
- Reduce the number of maintenance visits caused by network fault (incident or problem), by splitting regular maintenance visits into “proactive visits” and “regular visits”.
- Dispense only the people profiles needed to fix the anticipated fault
Medium-term predictions:
- Same as for short term plus
- In terms of planning the OSS network maintenance budget for the upcoming month
- Managing in a more effective way network operations contracts with Outsourcer(s) and “spare parts stock”
- Other business benefits (SLA penalties, Churn, etc..) reflected in the business case
Alarms
- Can anticipate actions in a call center and therefore improve customer experience
- Are “signals” that help predict incidents and problems
- Easier identification of problems (reoccurring incidents)
- Reducing false-positive alarms and associated time & efforts