Network Data Analytics Function is first out of the gate
Oracle this week announced Network Analytics Suite, a new portfolio of cloud-native network analytics solutions created to help Communication Service Providers (CSPs) gain insights around the performance and stability of their 5G network core. The first offering in the new suite, Network Data Analytics Function (NWDAF), is now available.
The new suite of solutions combines machine learning and artificial intelligence to help network operators more quickly identify and rectify network anomalies that can cause catastrophic failures, said Oracle.
The goal of the new suite, said Oracle SVP and GM Andrew Morawski, is to help CSPs achieve greater cost efficiency, improve quality of service and find new revenue streams.
“Our 5G Network Analytics Suite harnesses Oracle’s cloud and network experience into a powerful toolset that helps operators better manage and optimize their 5G networks,” said Morawski.
OC-NWDAF, as Oracle is calling it, is built to be cloud-native, said Oracle, aligning with Cloud Native Computing Foundation (CNCF) principles, and supporting 3GPP standards, including use cases identified in 3GPP Release 16 and Release 17.
“NWDAF as a core principle is defined by 3GPP standards to provide statistical and predictive analytics for 5G network function consumption,” said Oracle. It standardizes and streamlines how data is collected and how analytical insights drawn from the network are consumed. 3GPP has defined standard interfaces for 5G core network functions to deliver and receive the performance thresholds and analytics information, to and from the NWDAF.”
OC-NWDAF functionality is disaggregated into two logical functions: the 3GPP front-end, called NWDAF-FE (for front-end), and the Converged Analytics Platform for Communications (CAP4C). The NWDAF-FE provides a REST Application Programming Interface (API), which enables third-party apps “to request model execution of the standard use case categories and proprietary use cases that are supported,” said Oracle. It collects standards-defined data from the 5G network functions (NFs), app functions (AFs), and operations, administration and maintenance (OAM). Data is subsequently delivered to the analytics engine operating on the backend for more processing and analysis. The CAP4C can directly process data from the NWDAF-FE, or from other non-3GPP or proprietary sources using Apache Kafka, the open-source distributed event streaming platform.
“With the offering, CSPs can build automation into applications to monitor and audit software components. This supports anomaly detection to avoid events such as network function failure that could result in degradation of network quality. The offering can also support network monetization with data and insight sharing to third-party developers to create new services,” said the company.
Oracle sees the network automation that underpins technology like its new Network Analytics Suite as essential to the success of 5G. “There are three key words and they are automation, automation, automation,” said Anil Rao, Oracle’s senior director of product marketing and strategy, in a podcast with RCR Wireless News earlier this year.
“5G operations must be automation-native by design,” he continued. On the convergence of cloud, virtualization, disaggregation and decentralization of the network, “If you put all these things together. It’s fair to say that the networks that underpin 5G are going to look massively different. And it’s going to be immensely more complex to manage and operate on a day-to-day basis.”
