Access and Mobility Management Function

Description

The Access and Mobility Management Function (AMF) is a fundamental control plane network function (NF) within the 5G Core (5GC) architecture, defined by 3GPP from Release 15 onwards. It serves as the primary termination point for Non-Access Stratum (NAS) signaling, which is the protocol layer between the User Equipment (UE) and the core network, independent of the underlying radio access technology (e.g., NG-RAN, non-3GPP access). The AMF is responsible for a suite of procedures related to UE access and mobility, including initial registration, connection management (establishing and releasing the signaling connection), reachability management (including idle mode mobility and paging), and mobility management (tracking area updates, handovers). It authenticates the UE and authorizes its access to the network, acting as a security anchor by interacting with the Authentication Server Function (AUSF) and the Security Anchor Function (SEAF). A key architectural principle in 5G is the separation of the control plane functions for access/mobility (AMF) and session management (SMF). This decoupling allows the AMF to be selected based on UE location and mobility requirements, while a different SMF can be selected based on the data session's service requirements, enabling greater flexibility, scalability, and support for network slicing.

From a procedural standpoint, when a UE initiates registration with the 5G network, the Radio Access Network (RAN) routes the initial NAS message to an AMF instance. The AMF then orchestrates the UE authentication process. Upon successful authentication and registration, the AMF maintains the UE's context, which includes its identity (SUPI/SUCI), registration status, assigned temporary identifier (5G-GUTI), security context, and the serving SMF(s) for its active Protocol Data Unit (PDU) Sessions. For mobility, the AMF manages the UE's mobility within the 5G system, handling procedures like handovers (with the assistance of other functions) and tracking area updates. It also plays a central role in network slicing by being part of the slice selection process, ensuring the UE is served by the appropriate network slice instance based on its subscription and requested service.

The AMF interfaces with numerous other 5GC NFs. Its key interfaces include N1 (to the UE via the RAN), N2 (to the NG-RAN for control plane signaling), and N11 (to the SMF). It also communicates with the AUSF (N12), Unified Data Management (UDM) (N8), Network Slice Selection Function (NSSF) (N22), and Network Exposure Function (NEF) (N29), among others. This web of interfaces allows the AMF to act as a central hub, coordinating between the UE, the RAN, and other core network functions to manage the UE's access lifecycle. Its stateless design principle, where the UE context can be stored externally in a Unified Data Repository (UDR), enhances reliability and allows for efficient load balancing and redundancy across multiple AMF instances.

Purpose & Motivation

The AMF was created as part of the 5G Core's Service-Based Architecture (SBA) to address the limitations of previous core network architectures, specifically the 4G Evolved Packet Core (EPC). In the EPC, the Mobility Management Entity (MME) combined access, mobility, and session management signaling. This monolithic design limited flexibility, made network function scaling inefficient, and hindered the deployment of diverse services with different requirements on a shared infrastructure. The primary purpose of the AMF is to decouple access and mobility management from session management, a separation that is foundational to 5G's goals of network flexibility, service-based design, and support for network slicing.

By isolating the mobility and connection state management in the AMF, the 5GC can independently scale the AMF and the SMF based on different network loads (e.g., a massive number of idle IoT devices versus high-bandwidth data sessions). This separation directly enables efficient network slicing, as different slices can share a common AMF pool for basic connectivity while utilizing dedicated SMF instances tailored for specific slice performance characteristics (e.g., ultra-low latency, massive IoT). Furthermore, the AMF's design as a stateless function (with context stored externally) improves network resilience, simplifies disaster recovery, and enables more agile load balancing compared to the stateful MME. Its creation was motivated by the need for a core network that could support a vastly wider range of use cases—from enhanced mobile broadband to mission-critical communications and massive IoT—with the required agility, scalability, and efficiency that the previous generation's architecture could not provide.