Avatar

Older telecommunications networks weren’t built with 5G in mind. They were designed to provide voice, data, and basic connectivity services. But today, 5G is opening doors we never thought we would get a glimpse behind. Service providers are encountering new requirements, new customers, new use cases, and a myriad of connection types and in this journey, they are seeking the most effective ways to handle these increasing traffic demands.

Service isolation and stringent Service Level Agreements (SLAs) with high Quality of Service (QoS) requirements are forcing service providers to explore new solutions that allow them to keep their existing customers happy while also providing near-zero latency paths for other traffic flows. This was never more important than in today’s 5G networks, and it’s essential that service providers research and understand new technology that accommodates this growth.

In the past, transport networks designed to support 2G/3G/4G services were successful in deploying Unified Multi-Protocol Label Switching (MPLS) architecture, which is a solid solution. However, this approach also presents many complexities when deployed in large networks, including complexity in the control plane, the number of control plane protocols, and the high number of device-level configurations needed for a service spanning many domains.

Other 5G challenges include footprint expansion due to radio densification, incorporating the network or edge data center seamlessly into the transport network, use of Virtualized Network Functions (VNFs) for radio and mobile core functions, introduction of the Control Plane/User Plane Separation (CUPS) architecture, and implementation of network slicing.

Network Slicing in the Transport Layer

Network slicing is a key concept in 5G. It offers the ability to build segregated end-to-end networks serving different 5G use cases and customers. Network slicing is a flexible, scalable architecture enabling the multiplexing of virtualized, independent networks on the same physical infrastructure. When you partition mobile networks into a set of virtual resources, each “slice” can then be allocated for a different purpose.

An example would be to route latency-sensitive traffic such as gaming into one slice, while another slice is devoted to less latency-sensitive traffic like video streaming. Since each service resides is in its own slice, they run in isolation from one another. Other examples for 5G might be allocating a slice to an Internet of Things (IoT) domain, to an enterprise customer, or to a Mobile Virtual Network Operator (MNVO). It is a way to provide a variety of specialized SLAs and create a more intelligent and cost-effective network.

In 2017 the 3rd Generation Partnership Project (3GPP) standards body approved specifications for 5G communications, with the first set defining New Radio (NR) as it relates to Non-Standalone (NSA) operation (5G deployments on 4G systems). In 2018 the 3GPP set out specifications for 5G Standalone (SA) networks (greenfield 5G deployments).

The full 5G core system has three key elements driving network slicing:
• enhanced Mobile Broadband (eMBB): Originally introduced in the first phase of 5G NSA, eMBB provides greater data bandwidth and some improvements to latency for 4G LTE and 5G NR. This helps develop new use cases such as Artificial/Virtual Reality (A/VR) and video streaming.
• Ultra Reliable Low Latency Communications (URLLC): Provides increased bandwidth for 5G core deployment to provide end-to-end latency reduction.
• massive Machine Type Communications (mMTC): Developed to provide connectivity to a much larger set of devices like IoT sensors that transmit small data blocks through low bandwidth pipes.

Because network slicing is an end-to-end partitioning of network resources and functions, the slice refers to every aspect of the 5G architecture including radio, transport network, mobile core infrastructure, and the orchestration infrastructure necessary to manage and operate the slice.

One important requirement of network slicing is shared components between slices. Multiple slices might share the same mobile control plane, but each slice may also have its own user plane function. In order to make this possible, the requested mobile control plane function must be accessible to both slices, so the optimal solution is to use an IP layer with firewalling.

Network slicing must also meet the following requirements within the network infrastructure:
• Transport slice management: This is the ability to create, modify, and delete a 3GPP network slice, including any actions required on the transport layer. The slice application owner and the operator must be able to monitor the health and performance of the slice through pre-slice operations, administration, and maintenance (OAM) capabilities.
• Slice isolation: Each transport slice must be isolated from all other transport slices in order to meet stringent SLAs. This means the independent slice must meet proper performance, security, operation, and reliability levels.
• Resource reservation: This is the ability to reserve transport resources for a transport slice.
• Abstraction: It is essential that service providers have the capability to utilize resources required to model and build a transport infrastructure to meet the needs of a slice.

There are two types of network slicing: hard and soft. In each case they must meet the requirements we’ve discussed here, with the difference being the level of resource sharing between the slices, as well as the number of slices available on the network. A hard slice has dedicated resources, most notably bandwidth, that is not shared with other slices. Soft slicing, on the other hand, is more agile and flexible in nature, which means resources can be shared between slices while still maintaining SLA requirements, and the resources can be returned to the network when they are no longer needed.

Automation and Orchestration are Keys to Network Slicing

Network slicing automation and orchestration focuses on configuration, functionality, and operations across multiple domains. The goal of automating these functions is to reduce the need for human oversight in manually building, configuring, and maintaining slices.

Automation can be introduced into all domains of the network including the radio, Radio Access Network (RAN), transport, and mobile core. All these domains have new technologies in 5G that will need to be automated for effective implementation and management of network slicing. Our solutions are based on an open network architecture designed to support multi-vendor environments. Principle technologies include real-time analytics as well as model-driven telemetry, configuration APIs, and multi-vendor orchestration.

Our innovative network operating system, IOS XR, helps streamline both architecture and operations. This solution offers segment routing to optimize network utilization, Ethernet VPN (EVPN) to speed up service delivery, enhanced security, model-delivered telemetry for real-time visibility, modular packaging, and zero-touch provisioning to automate Day 0 and Day 1 provisioning of transport.

Service providers are quickly discovering the benefits of automation and orchestration of their network services as a way to deliver them faster and at a reduced cost in order to level the playing field with Over the Top (OTT) and cloud-service providers. They are finding this leads to faster time to market, simplified network operations requiring fewer skilled resources, and improved cash flow.

Automation and orchestration facilitate the consumption of new technology and time to market for new service delivery. They also offer the capability to discover and fix bugs more easily when introducing new technology. But the best part for service providers is the cost savings and increased revenue. In a 2019 study we conducted on the operational processes of major service providers that automated their network environment, the results showed a significant reduction in Operating Expenditures (OpEx). OpEx savings were observed in legacy, hybrid, and virtual services and infrastructure.

Learn more

Network slicing is essential for 5G service providers to deliver on stringent new SLAs while still producing quality service to their existing customer base, and we are excited to be at the forefront of delivering the next generation of technology solutions empowering 5G. We are always happy to help, so contact us with any questions you have on incorporating network slicing into your transport network and empowering it with automation and orchestration.

Also, be sure to check back with us for the next blog in our 5G series where we will have an in-depth look at our innovative packet core and cloud native solutions.



Authors

Laurent Desaunay

Technical Solutions Architect

Service Provider Sales