By unleashing the value of data, whether it be network data, customer data or data from other sources, operators can run their networks more efficiently and cost-effectively, and enhance customer experience. And in a virtualized environment, analyzing data isn’t just beneficial – it’s critical.
The Predictive Analytics for NFV Resource Management Catalyst proof of concept is setting out to prove the benefits of using predictive analytics to anticipate what may happen in the network. The goal is “to detect, monitor and act on unexpected network events, even if these events come from outside the network, for example from social media or IT,” says Samia Benrachi, Business Innovation Telco Architect at Atos, one of the Catalyst participants.
This holistic approach using multiple sources will provide greater ability to predict the needs of the network and provide a feedback loop. “We need real-time response and action in resource planning,” Benrachi says.
Increasing revenue and lowering costs
The Catalyst, which will be demonstrated at TM Forum Live! in Nice next week will also explore how to predict unusual usage patterns and will even be able to flag up the need for network upgrades or problems. The value proposition targets three areas: revenue generation, revenue protection and operational efficiency, and will focus on service schemas and predictive scenarios.
As operators deploy network functions virtualization and software-defined networking, adopt platform architectures, and move toward 5G, data analytics become even more important. Traditional capacity management will no longer work because old models tend to be deterministic and rely on resource hierarchy scenarios for planning. With increasingly complex service delivery chains, the need to be able to identify bottlenecks and respond immediately and without human intervention is critical.
The champion for this important Catalyst is French operator SFR, and participants include Atos, Dell and Ciena. One deliverable for the group will be to extend and deepen the existing metrics in the Business Metrics Specifications Model (GB935-A).
To learn more about predictive analytics, check out this TM Forum webinar on demand.
Watch Atos’ Samia Benrachi discuss this Catalyst project below, and if you’re headed to Nice, be sure to stop by the Catalyst Zone to see it action.