Making business sense with NWDAF

With 5G’s exponential data growth, leveraging MLOps and automation in conjunction with NWDAF allows organizations to proactively manage network performance, security, and user experience.

Data is becoming a crucial production element – data collection, storage, processing, and an understanding of the economic specifics of data are key strands that provide both challenges and opportunities. With data and analytics key to value creation, CSPs adopt standards-based analytics functions to drive actionable insights in their operations.

Network data analytics function (NWDAF) is an analytics function that provides operational intelligence, streaming, and data collection. Guavus and Sandvine are two vendors, among others, that offer full-featured and vendor-agnostic NWDAF capabilities that enable CSPs to place data and analytics at the heart of their innovation model. Also, NWDAF functionality is a stepping-stone for CSPs to build analytics functions that propel them forward in the digital economy. “In a digital economy, where applications and APIs rule supreme, CSPs seek to become data-driven organizations. They should apply analytics at multiple ‘stations’ of the network spanning core, transport, and edge locations,” says Don Alusha, Senior Analyst 5G Core & Edge Networks at ABI Research.

CSPs do not seek a one-time attempt at innovation but rather sustainable productivity growth as a source of competitive advantage. With 5G, cloud, and big data analytics, the industry has the external catalysts that can drive new growth and innovation. But CSPs, and the broader industry, must realize that growth will come from combining and rearranging the existing (cellular) technologies with big data and analytics solutions. New use cases need to be assessed on business impact and complexity in terms of skills required to deliver them. In addition to having the right tools and analytics functions, success for CSPs will come from how they use those tools. Orange, Telefonica, Vodafone, and Verizon are among the many CSPs investing to obtain the right know-how.

NWDAF, a standard created by 3GPP enhances 5G network performance by analyzing data from network functions. Earlier, in Release 15, 3GPP had laid emphasis on the 5G core (5GC), but with growing data, the need for data analytics has become seemingly important. 3GPP, as part of its Release 16, announced enhancements in terms of network automation through NWDAF, and made enhancements in terms of location service, network automation, service-based architecture, network slicing, and the same release goes on to discuss on non-public networks (NPN), and details out on providing support for cellular IoT. Further, in Release 17, the 3GPP details out on how to support multiple NWDAFs in a single network, which allow for multi-vendor domains to exist.

Using NWDAF, abnormal network behavior detection is the predictable user behavior, unusual network gets detected, operators are able to manage the network effectively and react to changing user demands, with zero human intervention. NWDAF introduces higher level of intelligence to the 5G network, enabling 5G networks to become smarter with means of real-time-based data management and analysis, which will enable the operator’s business in terms of the effective functioning of the network, and also help the operators to monetize it. Network analytics is new to operators but adopting it faster will optimize and enable better business for them.

Often, service providers face challenges to integrate proprietary solutions along with their own features set, posing a complete vendor lock-in, while the network is working in silos. Breaking away, NWDAF is designed to operate with an open framework that is capable of interacting with any network function (NF), enabling business to be more scalable and highly automated that will enable easier 5G deployment and management with enhanced better customer experience.

NWDAF creates enhancements to network automation with the application of analytics, and provides intelligent automation of the network. In order to provide enhanced rich end-user experience, standard interfaces from the service-based architecture (SBA) collects data by subscription or by means of a request model from other network functions and delivers analytics functions in the network for the purposes of automation or reporting, solving major custom interfaces or format challenges. NWDAF will stand out as it is not only a technical decision but a business decision as well.

The basic premise of NWDAF architecture is that it supports data management and analytics services. NWDAF ex­tracts data from micro-service functions residing inside the network core, operations, administration, and maintenance (OAM) systems, so that relevant data can be analyzed at scale. Some vendors have developed it further to analyze data residing in other domains of the network like RAN, application function layer as well. NWDAF helps in analyzing granular insights that will enhance network automa­tion and service orchestration, resulting in optimized network and increased service levels.

In the present context, there is a sea of data sitting across the core, transport, and edge domains, and that volume will grow exponentially alongside is the increased customer expectations (IoT-driven auto­nomous and smart services, fixed wireless access, enhanced mobile broadband), will grow accord­ingly through devices, locations, OTT services, and network topologies. Operators and enterprises will need to effectively unlock the potential of 5G services, and will need to have a good deep understanding of what customers are experiencing in real time.

The objective is clear that predictive analytics platforms needs to be well designed to examine data from 5G core networks and fuel insight-driven decision, providing a true case for a service assurance approach to 5G, which in turn will require analytics throughout every domain of the network. At present, the data that is being analyzed and applied is very marginal, and business needs to consider adopting NWDAF early.

NWDAF, powered with AI/ML, empowers closed-loop network operations and network anomalies, such as predictable user behavior and unusual network that is detected. Operators are able to manage the network effectively and react to changing user demands, with zero human intervention.

Business potential with NWDAF

• NWDAF coexists with other components of the organization’s analytics data analytics architecture;
• Reduced TCO as NWDAF can integrate and is capable of re-using already existing data analytics;
• NWDAF aligns and supports various other AIOps initiatives undertaken by the business;
• Enhanced user experience with analytics-as-a-service provides monetization opportunities;
• Can be adopted for 2G/3G/4G machine-learning net­work operations; and
• Goes beyond network analytics.

In a nutshell, dis­aggregation of NWDAF is enabling efficient data management and analytics will make a true case for network automation, and data sitting across domains can be analyzed, automated, and help business to perform better. NWDAF is a new threshold for analytics, going beyond just network analytics, but is designed to deliver end-to-end service assurance with cloud efficiency and optimising network resources. Network analytics is not new to many of the software and cloud companies, as they already are adept in handling big data and analytics; however, from the operators’ standpoint, analytics may be a bit new, and they may take some time to really tap into its potential to make better business sense.