Data Collection Application Function

Description

The Data Collection Application Function (DC-AF) is a core component within the 5G Service-Based Architecture (SBA), specifically part of the Management Data Analytics (MDA) framework defined by 3GPP. It acts as a centralized policy control point and orchestrator for data collection tasks across the 5G Core Network (5GC) and, in some deployments, the Radio Access Network (RAN). The DC-AF interfaces with other network functions, such as the Network Data Analytics Function (NWDAF), Network Exposure Function (NEF), and various data producers (e.g., AMF, SMF, UPF), using standardized service-based interfaces (e.g., Ndcaf_DataCollectionProvisioning, Ndcaf_DataReporting). Its primary role is to translate high-level analytics or monitoring requirements—often received from an operator's Operations Support System (OSS), Business Support System (BSS), or a consumer Application Function (AF)—into specific data collection jobs.

Architecturally, the DC-AF operates by receiving data collection requests, which specify what data to collect (e.g., specific KPIs, events), from whom (target Network Functions or UEs), under what conditions (triggers like thresholds or periodic intervals), and where to deliver the collected data (e.g., to an NWDAF for analysis or an external AF). It then provisions these collection policies to the relevant data producers via its southbound interfaces. These producers are equipped with Data Collection Clients that execute the policies, collect the specified data, and report it back to a designated Data Collection Consumer, which could be the DC-AF itself, an NWDAF, or another authorized entity. The DC-AF manages the lifecycle of these collection jobs, including activation, modification, suspension, and termination.

Key components within the DC-AF's operation include the Data Collection Provisioning Service, which handles the creation and management of collection jobs, and the Data Reporting Service, which may receive and forward collected data. It leverages the 3GPP's Common API Framework (CAPIF) for secure service exposure and discovery. The DC-AF is defined to support a wide range of data types, including performance measurements (e.g., latency, throughput, error rates), mobility events, session management events, and user equipment (UE) analytics data. It can collect data on a per-UE basis, per-network-slice basis, or aggregated across a group of UEs or network functions.

In the broader network ecosystem, the DC-AF plays a critical role in enabling data-driven automation and intelligence. By providing a standardized, automated mechanism for on-demand data gathering, it eliminates the need for manual, ad-hoc configuration of probes or logging systems on individual network functions. This reduces operational overhead and ensures consistency in the data collected for analytics purposes. The DC-AF is a foundational enabler for closed-loop automation, AI/ML-driven network optimization, and proactive service assurance in 5G Standalone (SA) networks.

Purpose & Motivation

The DC-AF was created to address the growing complexity and data-intensive nature of 5G networks. Previous generations (4G/LTE) relied heavily on proprietary, vendor-specific solutions and manual processes for collecting network performance and user experience data. Operators often deployed separate probe systems or relied on inconsistent logging from network elements, leading to fragmented data silos, high integration costs, and slow time-to-insight. This approach was insufficient for 5G's dynamic use cases—such as network slicing, ultra-reliable low-latency communication (URLLC), and massive IoT—which require real-time, granular, and correlated data from multiple network domains for effective service assurance and optimization.

The primary motivation for standardizing the DC-AF in 3GPP Release 17 was to provide a unified, open, and automated framework for data collection within the 5G Service-Based Architecture. It solves the problem of how to efficiently gather the right data from the right sources at the right time to feed advanced analytics engines like the NWDAF. By defining a standard API and policy model, the DC-AF enables interoperability between equipment from different vendors and allows third-party applications (via the NEF) to request data collection in a controlled manner. This fosters innovation in network analytics and enables new business models where operators can expose network insights to vertical industries.

Furthermore, the DC-AF addresses scalability and efficiency challenges. In a 5G network with millions of devices and numerous network slices, manually configuring data collection per slice or per service is impractical. The DC-AF's policy-driven automation allows operators to define collection templates and apply them dynamically based on network conditions or service requirements. This reduces operational expenditure (OPEX) and enables the network to self-optimize based on collected data, moving towards zero-touch network and service management (ZSM).