5G Core Network

Description

The 5G Core (5GC) network is the fundamental control and connectivity framework defined by 3GPP for 5G systems, succeeding the Evolved Packet Core (EPC). It is architected as a Service-Based Architecture (SBA) where network functions (NFs) are modular software entities that expose their capabilities as services via well-defined interfaces, primarily using HTTP/2 and JSON. This cloud-native design, leveraging concepts like statelessness, microservices, and containerization, allows for flexible deployment, scaling, and lifecycle management independent of hardware. The 5GC physically separates the User Plane (UP) and Control Plane (CP), enabling distributed UP functions (UPFs) to be deployed at the network edge for low-latency services while centralizing control functions.

Key functional components include the Access and Mobility Management Function (AMF), which handles connection and mobility management; the Session Management Function (SMF), responsible for session establishment, modification, and release; and the User Plane Function (UPF), which is the anchor point for data forwarding and packet routing, inspection, and QoS enforcement. Other critical functions are the Authentication Server Function (AUSF) and Unified Data Management (UDM) for security and subscription data, the Policy Control Function (PCF) for policy governance, and the Network Repository Function (NRF) for service discovery within the SBA. The Network Exposure Function (NEF) securely exposes network capabilities to external application functions.

The 5GC operates by establishing a Protocol Data Unit (PDU) Session, which is a logical connection between the User Equipment (UE) and a specific Data Network (DN), such as the internet or an enterprise network. During initial registration, the UE interacts with the AMF and AUSF/UDM for authentication. For session establishment, the SMF, in consultation with the PCF, selects a UPF and establishes the necessary N4 interface rules for traffic handling. User data packets then flow between the UE (via the Radio Access Network) and the DN through the UPF(s), with the SMF managing the session state and the AMF handling mobility events like handovers. This architecture supports concurrent access to multiple data networks and multiple PDU sessions of different types (e.g., IPv4, IPv6, Ethernet, Unstructured).

A cornerstone capability of the 5GC is native support for Network Slicing. It allows the creation of multiple logical, end-to-end networks on a common physical infrastructure, each tailored with specific characteristics (e.g., bandwidth, latency) for different service types like enhanced Mobile Broadband (eMBB), Ultra-Reliable Low-Latency Communications (URLLC), or massive IoT. The 5GC identifies a slice via the Single Network Slice Selection Assistance Information (S-NSSAI) and ensures a UE's PDU Session is associated with the correct slice instance, with dedicated AMF, SMF, and UPF resources as needed. Furthermore, the 5GC architecture integrates support for Edge Computing, enabling application functions to influence traffic routing (e.g., via Local Area Data Network or UPF selection) to meet stringent latency requirements.

Purpose & Motivation

The 5GC was created to address the limitations of the previous 4G Evolved Packet Core (EPC) and to meet the diverse and demanding requirements of 5G services as outlined by the IMT-2020 vision. The EPC, designed primarily for mobile broadband, was a monolithic, hardware-centric architecture with tight coupling between network functions, making it inflexible and costly to scale or innovate upon. The explosion of connected devices (IoT), the need for industrial automation with ultra-low latency, and the demand for immersive experiences like AR/VR required a more agile, efficient, and programmable core network.

Historically, each generation of mobile networks introduced a new core network (e.g., GSM's circuit-switched core, UMTS's packet-switched core, 4G's EPC). The shift to 5G presented an opportunity for a radical architectural redesign. The primary motivations were to achieve greater flexibility through software-based, cloud-native principles; to enable efficient support for a vast array of services through network slicing; and to reduce operational costs through automation and scalability. The 5GC solves these problems by decoupling software from hardware, separating the user and control planes for independent optimization, and introducing a service-based interface model that simplifies integration and enables faster deployment of new features. It is the foundational enabler for 5G to be more than just faster mobile broadband, transforming it into a platform for vertical industries and new business models.

Key Features

• Service-Based Architecture (SBA) with HTTP/2 interfaces
• Control and User Plane Separation (CUPS) for flexible deployment
• Native support for Network Slicing via S-NSSAI
• Integrated support for Edge Computing and Local Breakout
• Cloud-native design principles (stateless NFs, microservices)
• Concurrent multiple PDU Session types (IP, Ethernet, Unstructured)

Evolution Across Releases

Defining Specifications