Unveiling the Power of NetWarden’s Anomaly Detection

Anomaly Detection, a fundamental tool in network monitoring, has undergone significant enhancements within the NetWarden framework. With recent updates, this algorithm not only aligns seamlessly with the latest 5G technology but also maintains compatibility with legacy technologies like 2G, 3G, and 4G. Let’s delve into the key aspects of these advancements and how they enhance anomaly detection efficacy.

One pivotal improvement lies in the algorithm’s newfound generality. Previously, its performance varied across different Key Performance Indicators (KPIs). However, irrespective of KPI type or behavior, the algorithm now demonstrates a remarkable ability to learn and identify relevant anomalies. This adaptability is empowered by cutting-edge Machine Learning techniques, ensuring a comprehensive approach to anomaly detection.

Moreover, the algorithm’s enhanced capability to discern both individual cell behavior and group dynamics represents a significant leap forward. By considering the unique characteristics of each cell alongside its collective behavior within a group, NetWarden’s Anomaly Detection achieves unprecedented precision.

Let’s highlight some crucial aspects of this Anomaly Detection system:

• Automatic Detection: Abnormal behaviors are identified automatically, minimizing manual intervention.
• Efficiency: The system eliminates the need for laborious manual adjustments, such as “top 10” lists.
• Generality: Leveraging Machine Learning, the algorithm adapts to diverse KPI behaviors, ensuring robust anomaly detection across the board.
• Global Insight: By grouping cells with similar behavior, the system reduces the likelihood of false positives, for example in sensitive KPIs like Success Ratio (SR).
• Robustness: Seasonal behavior is meticulously accounted for, mitigating the risk of false negatives, especially in traffic-related KPIs.

Rodrigo Borges, Head of Data Science added, “We were able to achieve these upgrades thanks to Bwtech’s expertise in telecommunication and experience gained by developing and improving Anomaly Detection’s first version. Now, with focus on data quality and monitoring, our Data Science team keeps working on continuously improving the new version.”

Therefore, it is worth mentioning that Anomaly Detection enables the detection of anomalies in KPI time series, considering seasonalities and the individual behavior of a cell and the group to which the cell belongs.

In summary, NetWarden’s Anomaly Detection emerges as a potent tool in the realm of network optimization. Its newfound adaptability, efficiency, and precision promise to revolutionize anomaly detection, ushering in a new era of network monitoring excellence.

For more insights into NetWarden and its advanced capabilities, contact us at [email protected]

This project was developed with the support of FAPEMIG, Finep, and the Ministry of Science, Technology, Innovations, and Communications (MCTI), with resources from FNDCT.

ㅤ