Ingestion and Normalization
SONAR4 ingestion is optimized for high volumes of real-time network traffic with millisecond precision. Control data, forwarding/metering data, application/microservice data, etc. is continuously processed, curated and prepared for analysis.
Telemetry Stream Enrichment
SONAR4 curated telemetry is enriched with additional time-aligned, contextual metadata. Such enrichment enables precise anomaly detection and root-cause analysis. OSS operational markers, NFV service chain, NFVI and similar data adds additional context.
SONAR4 offers deep learning algorithms to
- Detect anomalies and outliers
- Identify QoS degradations and faults
- Perform fault correlation and root cause analysis
- Identify SLA violations
SONAR4 is microservices-based, so CSPs can select the specific features, protocols and time-sensitivity required.
Processed data is preserved in time-series databases, based on the specific timescale needs of the SDN Controllers and Orchestrators. Ad-hoc analysis can be performed within SONAR4 or third-party toolsets. New ad-hoc signatures can be loaded into SONAR4 for real-time processing.
Any detected anomalies, faults or SLA violations can alert the SDN Controller or NMS systems to trigger remedial actions. SONAR4’s Policy Engine can trigger additional testing before alerts are raised (e.g. activate smart vProbe and observe S1-U traffic for target subscribers for a given time).
SONAR4’s Operations Portal provides easy-to use visual aids to CSPs’ network operations personnel. The processed data can be queried and analyzed by either SONAR4’s Analytics Explorer or third-party analysis and visualization tools. Data can easily be exported to data lakes and northbound analytics applications.