To deliver the new services promised by 5G, your existing 4G core will not be enough.

David Nowoswiat, Nokia

August 9, 2018

4 Min Read
Why Do You Need a 5G Cloud-Native Core?

Changing the core network is never done lightly and with 5G it will be no different. There's a lot at stake with evolution to a new core, which is why some communications service providers (CSPs) are apprehensive about evolving from their current Evolved Packet Core (EPC), even as they add 5G New Radio (NR).

CSPs understandably want to accelerate standardization of the 5G non-standalone (NSA) radio architecture with EPC (evolved Packet Core) -- it can use the existing 4G EPC as the core and 4G/LTE RAN for network coverage, allowing them to add 5G NR to address network capacity constraints with minimal changes to networks. Nonetheless, an NSA network isn't a complete 5G network until the core supports the new 5G core (5GC) functions. Upgrading an existing EPC with new 5G NR to expand existing mobile broadband services may be a short-term solution, but it won't meet long-term objectives.

Upgrading your current EPC won't be enough
CSPs cannot economically or technically support the many devices connecting to the network and meet 5G services requirements by upgrading an existing EPC. To achieve the necessary scalability, flexibility and performance requires a cloud-native, 5G service-based core architecture that includes separate user and control planes (CUPS) and high-bit rate provisioning.

With 4G, a cloud-native core architecture is preferred, but with 5G, cloud-native design is essential. Only by redesigning the software architecture and core functions using cloud-native principles and IT methodologies, can CSPs reduce time-to-market for services.

Services Based Architecture
The new 5GC introduces a new Services Based Architecture (SBA) where the 5GC control plane is based on services exposed by network functions (NFs) using new service-based interfaces (SBIs). Once a 5GC function registers its services with the new 5G Core Network Functions Repository Function (NRF), it exposes services to all authorized consumers, rather than defining a new point-to-point interface and procedures between the two functions as an EPC requires. This offers CSPs greater flexibility and more efficiency by decoupling the service consumer from the service producer.

Improved session management
Session management is improved with a new session and service continuity mode that supports a "make before break" option and the relocation of core network functions while maintaining seamless end-user services. This is essential for ultra-reliable low-latency communications use cases with moving UEs (e.g., V2X). It also supports concurrent local and central access to a data network and Multi-access Edge Computing (MEC), where the application at the edge data center can influence traffic routing to improve performance.

Flow-based QoS framework
The 5GC supports a more granular, flow-based QoS framework compared to the static bearer based QoS of EPC, which means QoS can be assured for each application. For real-time services such as AR/VR, this new QoS framework will ensure the best user experience.

Flexible end-to-end network slicing
End-to-end network slicing includes the RAN, the core and the transport network. Network resources can be dedicated per slice, allowing end-to-end service differentiation. Both vertical and horizontal slicing can be use the same service slice, enabling different slices and SLAs per tenant. For example, a dedicated slice can be allocated to separate use cases or services such as utilities, healthcare and automotive. Slices in the same vertical can serve different manufacturers, each having its own SLA. Also, end-user devices can simultaneously access more than one slice.

Access agnostic for seamless mobility
The 5GC is also access agnostic, with unified registration, authentication, session, mobility and policy management for all access types. Common interfaces for 3GPP and non-3GPP access enable seamless mobility. This will allow CSPs to consolidate all their services on the new core. Operators should be able to deploy separate 5G network slices for fixed-only, mobile-only and converged services.

Combining these features allows economic deployment of services at scale for new applications and verticals. At Nokia Bell Labs, we are exploring with BT Labs how a 5G core could increase the value of a converged network. At 5G World, we demonstrated how fixed and mobile access networks using a 5G core can support telepresence-like applications with resilient end-to-end low latency.

In the near term, CSPs will likely deploy a 5G NSA network (most likely option 3) to roll out initial services and get familiar with the new technology. This approach minimizes disruption, but the full benefits of 5G depend on a 5G core. CSPs that adopt a full 5G end-to-end network early can offer the scalability, performance and flexibility needed to economically deliver a broad range of services. The race for 5G leadership is on.

— David Nowoswiat, Snr Product Marketing Mngr, Nokia Corp. (NYSE: NOK)

This blog is sponsored by Nokia.

About the Author(s)

David Nowoswiat

Nokia

Dave is a senior product marketing manager at Nokia with over 25 years of telecom industry experience. Lately he's been doing his part to ensure cloud and 5G next gen packet core is in the limelight. In his spare time Dave plays sports, runs the odd marathon and enjoys good craft beer - not necessarily in that order.

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like