NWDAF is defined by the 3GPP as a standard for using network data analytics to drive the automation of 5G network operations. Defining the standard involves identifying specific use cases. Then, for each use case, specifying the input data that NWDAF needs to process to produce the required analytics outputs. Those outputs are consumed by 5G Core functions to automatically perform service orchestration and network operations tasks.
In this post, let’s look at the set of NWDAF use cases considered by the 3GPP, and why these are critical for enabling mobile network operators (MNOs) to operate complex 5G networks at scale.
NWDAF Processing Model
NWDAF continuously collects network-related data from multiple sources, including standard data types from Network Functions (NFs) in the 5G Core, as well as non-standard data from sources including Application Functions (AFs), OAM layer functions and non-standard data sources such as network probes.
NWDAF performs statistical and/or predictive analytics on collected data, applying AI/ML algorithms to generate the specified analytics outputs, which are consumed by NFs, AFs and functions in the OAM layer.
3GPP Release 16 NWDAF Use Cases
As MNOs scale out 5G Standalone networks, the adoption of NWDAF-based network analytics will be guided by the operational requirements for achieving business objectives.
3GPP Release 16, which was frozen in 2020, identifies the following NWDAF use cases, which each address particular operational requirements:
- Network slice instance load level computation and prediction
- Service experience computation and prediction for an application or network slice
- Load analytics information and prediction for specific NF instance(s) or types
- Application service experience computation and prediction
- Network performance computation and prediction
- User equipment (UE) mobility analytics and expected behavior prediction
- UE communication analytics and pattern prediction
- UE abnormal behavior or anomaly detection
- Current and predicted congestion for a specific location, device or group of devices
- Quality of service (QoS) sustainability – reporting and predicting QoS change
While these use cases encompass a broad range of 5G operational requirements, this list is only a starting point. The 3GPP will be defining more use cases in future releases. But what considerations drove the 3GPP to focus on this initial set?
NWDAF Operational Intelligence
5G network data analytics is predominantly concerned with generating operational intelligence related to the current and future state of:
- Network Conditions
- Device Behavior
- Service Experience
Release 16 NWDAF use cases provide real-time insights into these critical aspects of network and service operations. At a high level, the objectives are straightforward. Operators need to know – in real time – how users and applications are impacting the 5G network, or how adverse network conditions or unusual device behavior is impacting users and applications. Regardless, operators need to constantly monitor service experience metrics for each type of 5G service.
Of course, the ultimate goal is network automation, but each Release 16 NWDAF use case is also applicable to operator-centric 5G network operations, in which humans evaluate the operational intelligence generated by NWDAF and make decisions to manually initiate service orchestration and network operations workflows.
Here is a quick run-down of the types of monitoring and analytics that NWDAF is concerned with in each category of operational intelligence. Note that NWDAF calls for using statistical analytics to determine what is currently happening in the 5G network (or what has happened in the past) and predictive analytics to determine what is expected to happen in the future, based on current or historical trends. Predictive analytics will play a key role in alerting operators to looming problems that could result in service degradation or even possible outages.
Real-time network monitoring is typically per slice and involves generating performance metrics such as throughput, latency, and connection setups, measuring network load and detecting congestion or other types of network performance anomalies.
Both human users and machines will be connected to the 5G network. 5G operations require monitoring device connectivity, mobility and communications patterns. Abnormal device usage or unanticipated behavior by human users or smart machines could have a negative impact on network performance.
MNOs need to measure and track service experience to ensure that the 5G network is meeting user expectations as well as stringent performance and reliability service-level agreements (SLAs). NWDAF generated metrics will measure service experience by user, application type, device group or geographic location. The specification calls for tracking QoS sustainability to monitor the current quality of service and predict future changes.
The More Things Change …
Looking at the overarching goal of 5G network data analytics through this lens, we can say plus ça change, plus c’est la même chose – the more things change, the more they stay the same. Managing today’s 4G/LTE networks involves the deployment of various monitoring and analytics tools that track network conditions, device behavior and service experience. Managing 5G networks requires the same type of information and insights. Fortunately, the network itself is able to provide much of the telemetry and other measures needed, allowing AI/ML to augment the abilities of human operators, enabling MNOs to master the overwhelming complexity of massive-scale 5G networks.
Read more on 5G Analytics & NWDAF:
- Guavus 5G-IQ NWDAF
- How Do You Standardize 5G Network Automation?
- 5G Analytics Standards: Who Needs ‘Em?
- Will 5G Complexity Overwhelm MNOs?
- NWDAF: Automating the 5G Network with Machine Learning & Data Analytics
- 5G Rollout Challenges & Analytics-Driven Remedies for MNOs
To learn more about Guavus’ 5G NWDAF approach, schedule a meeting with Guavus today.
About the Author
|Andrew Colby is Head of 5G Strategy and Product Management at Guavus, a pioneer in AI-based analytics for communications service providers.
As a member of the Guavus Office of the CTO, Andrew leads initiatives with customers to identify ways to apply analytics to improve and transform their operations and customer experience.
He has worked in the areas of telecom and IP networking, operational support systems, and data analytics, for more than 30 years.
Image attribution: iStock