Understanding 5G Service-Based Architecture
2020-06-30 | 5 min read
The migration to 5G represents a vast revenue opportunity for players throughout the wireless communications ecosystem, including mobile network operators (MNOs). But the massive number of use cases, the incorporation of technologies such as network slicing, and 5G/LTE coexistence requirements pose significant challenges. Much of the technology involved is fundamentally new to MNOs.
Each of the primary 5G usage scenarios specified in the International Mobile Telecommunications-2020 (IMT-2020) standard — including enhanced mobile broadband (eMBB), ultra-reliable low-latency communications (URLLC), and massive machine-type communications (mMTC) — require the implementation of a service-based architecture (SBA). As the label implies, a service-based architecture emphasizes services as the primary architecture component used to put in place and execute functions.
In an SBA — as defined by the 3rd Generation Partnership Project (3GPP) — a set of interconnected network functions (NFs) deliver the control plane functionality and common data repositories of a 5G network. SBA and other concepts applied within the 5G core (5GC) like network slicing, control user plane separation, and edge computing, represent the fundamentals of the commercial success of 5G, enabling the delivery of various services across many vertical industries.
5GC NFs include:
• Authentication server function (AUSF), which authenticates UEs and stores authentication keys.
• Access and mobility management function (AMF), which manages UE registration and authentication (via AUSF) and identification (via unified data management) and mobility. The AMF also terminates non-access stratum (NAS) signaling.
• Network exposure function (NEF) that exposes capabilities and events. It stores the received information as structured data and exposes it to other NFs.
• Network repository function (NRF), which provides service discovery between individual NFs, maintaining profiles of NFs and their functions.
• Network slice selection function (NSSF), which selects the set of network slice instances serving the UE and determines which AMF to use.
• Policy control function (PCF), which provides policy rules to control plane functions.
• Session management function (SMF) that establishes and manages sessions. It also selects and controls the user plane function (UPF) and handles paging.
• Unified data management (UDM) that stores subscriber data and profiles. It generates the authentication vector.
• UPF is responsible for packet handling and forwarding, mobility anchor, and IP anchor towards the internet. It performs quality of service (QoS) enforcement.
Figure 1 shows NFs within the control plane enabling other authorized NFs to access their services.
Previous generations of wireless communications technology relied on NFs connecting over standard interfaces. However, with the migration to a cloud infrastructure, the so-called point-to-point model no longer makes sense. A point-to-point architecture contains many unique interfaces between the various network elements, creating dependencies between them and making it difficult to make changes to the existing network.
The transition to SBA separates the end-user service from the underlying network, making for a more agile network. The SBA also borrows many aspects from software-defined network technologies. For a true SBA, NF services must be self-contained and reusable. NF services should also use management schemes autonomously, separate from other services offered by the same NF. This new environment pushes companies to design cloud-native 5GC functions or virtual network functions.
For more information on 5G and Keysight's 5G portfolio, please visit the following webpages: