Extracting And Correlating Relevant QoE And KPI Data, Using Analytics

A conversation with Anil Singh, Director of Product Management, RAN

Operators have access to extremely granular data on their networks and their subscribers. To benefit from this data while avoiding to be overwhelmed by it, they need first to identify the level at which data has to be analyzed and stored to optimize traffic management effectively, and second to analyze it in real time to take preventive or corrective action. This is not trivial as the appropriate level of analysis depends on what the operator is trying to achieve and is the cause of the performance issue the operator is facing. Guavus works in this area using an analyze-first, save-later approach.

Anil Singh, Director of Product Management RAN, told us about how Guavus filters KPIs and other performance metrics using smart filtering: “With smart filtering, we can identify whether the data is relevant or irrelevant. By irrelevant, I mean that 90% of the data today that is generally used for reporting purposes only. The data used for reporting has limited use in tactical operations management, and so you don’t need to store that data.” When the system detects an anomaly in the network, “it must adapt and learn. The system needs to know that something went wrong, and now it needs to start storing more granular data,” Anil told us.

Show more