Network Data Analytics Function (NWDAF) | RADCOM ACE

At the heart of RADCOM’s NWDAF solution is RADCOM’s next-generation solution for 5G assurance – RADCOM ACE – that identifies the network slice instance and creates the slice utilization KPI’s that are provided to the PCF and NSSF per network slice instance.

RADCOM’s NWDAF allows an NF consumer to subscribe to and unsubscribe from periodic notifications of the KPIs and to subscribe to and unsubscribe to notifications when a threshold is exceeded.

QoS flows in different network slices

The PCF takes the input from RADCOM’s NWDAF solution to assign more resources or steer traffic policies, which helps the operator run their network slices more dynamically, while the NSSF takes the load level information provided by RADCOM’s NWDAF for slice selection. The user plane and control plane transactions are associated with the SST, and SD and KPIs are aggregated by the SST and SD (exposed via the 3GPP standard interfaces).

RADCOM’s NWDAF supported services:

  • N23 interface: a reference point between PCF (Policy Control Function) and the NWDAF
  • N34 interface: a reference point between NSSF (Network Slice Selection Function) and the NWDAF
  • Nnwdaf_EventsSubscription which enables the NF service consumers to subscribe/unsubscribe for network slice specific congestion events notification from the NWDAF
  • Nnwdaf_AnalyticsInfo which allows the NF to service consumers to request and acquire analytics from the NWDAF

RADCOM’s NWDAF provides operators with the ability to capture data from both non-SBI interfaces (N1, N2, N3, N4, N6, N9) and SBI interfaces (N5, N7, N8, N10, N11, N12, N13, N14, N15) so that as 5G services roll out, RADCOM’s NWDAF ensures a smooth transition to the new core architecture; delivering a central point for network analytics

An enhanced NWDAF | RADCOM ACE

As well as serving as an NWDAF, RADCOM ACE also offers an efficient and cost-effective, containerized service assurance solution. This provides end-to-end network troubleshooting, as well as complete service and customer experience visibility— empowering the operator with an enhanced, more agile NWDAF solution.


End-to-end call and session tracing and troubleshooting

Providing operators with a state-of-the-art, modern troubleshooting application- layer (call/session tracing and packet analysis) that utilizes the latest web-client principles of cloud-based applications for collaboration and ease of use, and industry-leading for tracing and packet analysis applications.

Network analytics and smart alarming

It is offering network analysis and alarming, which is interoperable with the troubleshooting applications, allowing for a seamless drill down from one application to another—for example, performing a drill from a KPI Reporting into session information and from there into packet information.

This can also work in reverse by way of a “drill up” from a specific example to a broader view of the associated data, e.g., starting from a packet and viewing the entire session, which includes that particular packet.

Automation with Artificial Intelligence (AI) and Machine Learning (ML)

RADCOM’s NWDAF with AI/ML capabilities continually collects network data from the NFs and provides real-time analytics back to the NFs and the operator’s BSS/OSS systems, delivering continuous network analysis to help proactively manage the 5G network, essential for advanced 5G use cases.

Network slices will be monitored automatically for performance and resources adjusted to make sure the agreed SLA is delivered. AI and ML will be continually utilized to assist zero-touch slice management by forecasting resource utilization trends and proactively improving/configuring the network resources.

Advanced ML algorithms will utilize the information collected by the NWDAF for tasks such as mobility prediction and optimization, anomaly detection, predictive QoS, and data correlation.

Some of the use cases for future 5G standard releases and enhanced network automation using ML are:

    • Network congestion data – current and predicted for a specific location
    • NWDAF analytics exposure to applications, for example in Smart City applications such as alleviating urban traffic congestion
    • NWDAF-assisted predictable network performance
    • QoS sustainability (which requires predicting QoS changes)
    • UE abnormal behavior/anomaly detection
    • UE communication pattern prediction
    • UE expected behavior prediction

Further reading

Skip to content