This Post presents the default 5G architecture with new and reutilized elements, connected and cooperative networks, aggregated services, and key functions of evolved networks.
Architecture Options for 5G
At present, there are various options for adding a 5G network to the infrastructure depicted in Picture 1.2. 3GPP has considered the final architecture and the respective
Picture 1 High-level snapshot of the mobile network evolution and the supporting core elements.
Picture 1.2 The key elements of 2G, 3G, and 4G network architecture prior to 5G deployment.
evolution plan leading to the NR and core architecture. The summary of these options is presented below. The development path of a 5G radio access network (RAN), as well as a core network, requires a phased approach. There will be two parallel versions of the radio technologies:
- LTE as defined in 3GPP Release 15, referred to as evolved LTE (eLTE), as discussed in 3GPP System
- Native fifth generation’s New Radio (NR).
Furthermore, there will be two core network variants during the initial phase of 5G system (which is also referred to as 5GS):
- Evolved Packet Core (EPC) as defined in 4G System Architecture Evolution (SAE) technical specifications
- Next-Generation Core Network (NGCN).
The deployment strategies of these variants lead to 12 scenarios.
Picture 1.3 The high-level principle for the architecture of the LTE/EPC and 5G new radio (NR)/next-generation core network (NGCN). The 5G system provides the possibility to virtualize the network functions (NF), which is presented at a conceptual level in this format. When 5G is set to interwork via 4G EPC, the S-GW and P-GW are further subdivided into separate functional elements, each taking care of user and control planes.
The scenarios representing the standalone LTE/EPC and standalone NR/NGCN architectures are depicted in Picture 1.3
As an additional solution in the interim phase of 4G EPC support for the initial 5G deployments, the Serving Gateway (S-GW) and Proxy Gateway (P-GW) functions have been broken into two parts: user plane (UP) and control plane (CP).
Picture 1.3 depicts some of the most important new 5G-specific elements:
- AMF refers to the Access and Mobility Management Function.
- UPF refers to the User Plane Function.
- SMF refers to the Session Management Function.
- NF refers to a network function. 5G network has a number of these network functions for specialized tasks, in a virtualized environment.
The philosophy of the 5G is remarkably different from the previous generations due to the virtualization of the network functions. This results in new, more optimized ways of utilizing the resources in a dynamic manner upon the need, via network slices. Network slice refers to a logical network that provides specific network capabilities and network characteristics.
The MNO can have a number of these network slices that can be configured each optimizing them to certain use cases and environments such as broadband communications, critical communications and massive Internet of Things (mioT) communications. Equally, the essential parameters of each slice can be adjusted separately, including the level of security.
As soon as the first 5G services are deployed, there will be only a small number of 5G-capable terminals, and the NR coverage will be only limited.
For this phase, the solution is to deploy the 5G radio coverage gradually in such a way that the NR elements are connected directly to the LTE radio elements as depicted in Picture 1.4.
Another option is to provide both interfaces New Radio-Long-Term Evolution (NR-LTE) and NR-EPC for relying on the legacy core network of 4G either via the LTE radio or by connecting the 5G NR directly to 4G EPC.
The first set of technical specifications of 5G includes the concept for NSA (non-standalone) architecture, which will be the option for dual connectivity (DC). The DC itself is not new anymore as such as it was introduced in L’I’E 3GPP Release 12.
Where carrier aggregation (CA) makes it possible to serve different carriers by the same evolved NodeB (eNB), the DC serves the carriers by different backhauls, which refers to separate eNB elements, or combined eNB and 5G NodeB (gNB) elements.
DC, as depicted in Picture 1.5, refers to a device that is simultaneously connected to multiple NG-RAN nodes, thus forming multi-Radio Access Technology
Picture 1.4 The 4G EPC develops further for better supporting 5G connectivity, by dividing the S-GW
and P-GW functions into user and control planes.
Picture 1.5 5G Multi-DAT Dual Connectivity
Dual Connectivity (MR-DC) with 5G Core network as defined in 3GPP TS 37.340. There are two possible scenarios for DC:
- Next Generation Radio Access Network Evolved Universal Terrestrial Radio Access-New Radio Dual Connectivity (NGEN-DC). In this scenario, a device is connected to a Next-Generation evolved NodeB (ng-eNB), which assumes the role of a master node (MN), and to a gNB, which acts as a secondary node (SN). The ng-eNB connects to the 5G Core, while the gNB connects to the ng-eNB via the Xn interface.
- New Radio-Evolved Universal Terrestrial Radio Access Dual Connectivity (NE-DC). In this case, the device connects to a gNB, which assumes the role of an MN and an ng-eNB being an SN. The gNB connects to 5G Core and the ng-eNB connects to the gNB via the Xn interface.
Gradual Deployment of 5G
For the fluent introduction of the 5G services, 3GPP has designed a phased approach for the deployment of the 5G. It means that the 5G operators may select the optimal network strategy along with the route toward full 5G services as they develop based on practical equipment availability and further development of the 5G technical specifications as of Release 15 (referred to as the first phase of 5G), and Release 16 (second phase of 5G).
Picture 1.6 presents the 5G standalone scenario, which is called options 1 and 2 for 4G and 5G systems, respectively. Picture 1.6 also depicts the intermediate options paving the way for the fully native 5G standalone option 2:
- Option 3. The non-standalone scenario for the initial phase of the 5G deployments. The 5G terminal relies on both LIE and NR radio networks and solely on the 4G core.
Picture 1.6 The most concrete 5G deployment options summarized.
- Option 4. The evolution path of the 5G can lead to the intermediate step of non-standalone, NR-assisted architecture that relies on the 5G core.
- Option 5. As soon as there is Release IS LTE available, it can serve the further evolved LTE-Pro terminals while the very first NR serves the 5G-capable terminals separately. This option has been deprioritized.
- Option 6. The 5G terminal can be connected via booth 4G and 5G radio networks, while the 5G core takes care of the connectivity. This option provides the customers with the widest options to fully exploit the capabilities of the 5G, still providing the legacy connectivity via 4G. In this scenario, the 4G would take care of the major part of signaling while 5G radio delivers the data. This option can be compared, thus, with the principle of the carrier aggregation. This option has been deprioritized.
- Option 7. The 4G core can also take care of the 4G and 5G terminals in the intermediate stage.
The 5G network architecture can be depicted by using reference point representation or service-based interface representation. Picture 1.7 depicts the overall architecture and interfaces of the 3GPP 5G network in the traditional reference point format. As can be seen in the figure, the new terminology for the 5G core network is 5GC, whereas the NR system is referred to as NG-RAN.
Picture 1.7 The Overall 5G architecture
The logical interfaces of the 5G architecture are referred to as NG, Xn, and El. The complete description of these interfaces can be found in the following specifications:
- 3GPP TS 38.410 defines the general aspects and principles of the NG interface.
- 3GPP TS 38.420 defines the general aspects and principles of the Xn interface.
- 3GPP TS 38.470 defines the general aspects and principles of the Fl interface.
5G System Architecture
the 3GPP technical specification TS 23501 describes the 5G system architecture. This section presents some of the key architecture models referencing technical specifications.
The interaction between the 5G network functions can be represented in two ways; via the traditional reference point model or service-based model. The service-based representation refers to the network functions within the control plane enabling other network functions to access their services.
The reference point representation, in turn, depicts the interaction between the NF services in the network functions.
NFs have thus both functional behavior and interface. Each NF can be deployed as dedicated hardware forming the network element, or as a software instance that is relying on hardware. Furthermore, the NF can also be implemented as a virtualized function instantiated on a relevant platform such as in a cloud-based infrastructure.
Architecture of 5G
The service-based architecture model provides a highly modular and reusable approach for the 5G network deployments. It thus allows MNOs to deploy the 5G gradually, benefiting from the latest advancements of the virtualization concept. The service-based architecture contains the core network functions by interconnecting them simply
with the rest of the system. Picture 1.8 presents the visual model for depicting the service-based architecture, combined with the reference point-based architecture elements. the latter shows the interactions between network functions to inform about the system-level functionality, as well as inter-PLMN interconnection via different network functions.
Picture 1.8 Non-roaming 5G system architecture, presented via service-based interfaces.