The basics of network slicing and the benefits it offers from a telecom network perspective.

The edge, in this case, would be the telecommunications network edge. In the age of 5G, telecoms are moving from pre-packaged physical network appliances to a disaggregated stack running software network functions on different flavors of hardware.

The 5G network can be broken up into two main components — radio and core. It is the 5G core that controls the network and, therefore, becomes a key enabler for vertical market applications. One of the core features is network slicing.

This blog post will provide an overview of network slicing and some of the use cases at the edge.

Please make sure to check out all the installments in this series of blog posts on edge computing:

• Part 1: “Cloud at the edge”
• Part 2: “Rounding out the edges”
• Part 3: “Architecting at the edge”
• Part 4: “DevOps at the edge”
• Part 5: “Policies at the edge”
• Part 6: “Models Deployed at the edge”
• Part 7: “Security at the edge”
• Part 8: “Analytics at the edge”
• Part 9: “5G at the edge”
• Part 10: “Clusters at the edge”
• Part 11: “Automation at the edge”
• Part 12: “Network slicing at the edge”
• Part 13: “Data at the edge”
• Part 14: “Architectural decisions at the edge”
• Part 15: “GitOps at the edge”
• Part 17: “Storage services at the edge”
• Part 18: “Cloud services at the edge”
• Part 19: “Distributed cloud: Empowerment at the edge”
• Part 20: “Data sovereignty at the edge”
• Part 21: “Solutioning at the edge”
• Part 22: “Connected products at the edge”
• Part 23: “Foundational models at the edge”

What is networking slicing?

Network slicing is not new; it is a way of creating multiple unique logical and virtualized networks on top of a shared infrastructure using software-defined networking (SDN). As we alluded to in earlier blogs, 5G technology has accelerated the use of SDN and network functions virtualization (NFV) — both of which help in the quick creation of network slices. Thus, applying the same principles of virtualization to the radio access networks (RANs), a network operator can physically segregate traffic through slices.

Network slices can support a specific application, service, set of users or network and can span multiple network domains, including access, core and transport. They can also be deployed across multiple operators. This means that each logical network is designed to serve a defined business purpose and comprises of all the required network resources — configured and connected end-to-end.

One of the most important features of 5G, network slices can be dynamically created and programmed to provide end customers with their own individual mobile network and subscriber management, which can be continually updated to meet their evolving needs. From ultra-high bandwidth to ultra-low latency, the use cases identified for 5G and network slicing span diverse delivery needs and fall into three major categories. The 3GPP (3rd Generation Partnership Project) Release 15 provided details about these three Slice/Service types (SSTs):

• SST 1 – Extreme (or enhanced) Mobile Broadband (eMBB): Meant for video-centric applications that consume a lot of bandwidth and generate the most traffic on the mobile network.
• SST 2 – Ultra-reliable Low-Latency Communications (urLLC): This will allow for things like remote surgery or vehicle-to-X (v2x) communications and require Mobile Network Operators (MNOs) to have mobile edge computing capacity in place.
• SST 3 – Massive Machine-Type Communications (mMTC): While commonly known as Internet of Things (IoT), this network slice services at a much larger scale, with billions of devices being connected to the network. These devices will generate far less traffic than eMBB applications, but there will be many magnitudes more of them.

Service concurrency is enabled by network slicing. There is an intrinsic effect where each service has intrinsic characteristics: a micro effect wherein change in one slice does not affect the adjacent slice, and a macro effect whereby physical resources are moved to another slice based on demand. NFV is what provides the scalability, flexibility and isolation.

It is important to understand that network-level isolation — where vertical customers do not share network function or resources with the other customers — in some cases is not considered to be a fundamental requirement and is not addressed by network slicing. Consequently, network slicing assumes that only a small number of slices will be required in a national network and that they will be shared by enterprises having similar needs. 

Private networks

Network slicing without network isolation may be acceptable in specific use cases that require national network coverage (like connected cars), but many enterprises do not want to share any hardware or software infrastructure with other enterprises, let alone their competitors. Hence the emergence of Private LTE and Private 5G.

Network slicing should not be confused with Private LTE/5G as they cater to different applications. Private networks that we see in manufacturing plants, airports and ports, utility production, distribution centers, etc. require and use Dedicated Core Networks. These Dedicated Core Networks provide the enterprise with full network isolation, bringing greater control, reliability and deterministic quality because they are not shared with other customers.

Network slicing drivers

Network slicing caters very well to applications like nationwide fleet management, providing a reliable and controlled service wherever the device is located. Some of the drivers of network slicing are depicted in Figure 2.

5G has facilitated virtual network functions (VNF), especially network slicing, to fully realize the flexibility that virtualization promises. However, the underlying software applications for network functions must be architected to support any infrastructure and fully automate deployments and lifecycle events like service creation, transparent software upgrades, dynamic scalability and even recovery:

Network slicing questions

When and how should the network slice be created? Network operators have to answer these and other pertinent questions, including the following:

• How do you create the slice? (Per service, per vertical, per app, per QoS.)
• Where do you create the slice? (Cell, RDC, local DC, core.)
• How many physical resources will a slice require? (Location/size of slice.)
• What are the latency targets?
• How will you allow for scaling? (Allocation of physical resources to virtual slice.)

Steps in the lifecycle of a network slice

The steps in the lifecycle of a network slice include preparation, commissioning, operation and decommissioning:

• Preparation: Evaluate network slice requirements. Design/select a slice topology. Create and onboard a network slice template.
• Commission: Create the Network Slice Instance (NSI) and the Network Slice Subnet Instance (NSSI) that manage the slice requirements.
• Operation: This step includes NSI/NSSI activation, needed modifications and deactivation.
• Decommission: Whenever possible, terminate the instances.

Slice requests are decomposed across each layer of the telecom cloud stack, including the network control layer, the logical network layer, and the edge connectivity layer. The graphic below shows the calculations that go into deciding how to slice the network while evaluating certain policies and criteria at each step:

5G network slicing is an order-driven process, and products like the IBM Cloud Pak® for Network Automation provide a single graphical interface that network operators can use to perform the complex task of network slicing and manage the slices. It provides intent-driven orchestration, extreme automation and optimization features. Such tools let operators monitor the status of the slices and SLA metrics, and give them the ability to quickly address any alarms or modifications.

Wrap-up

5G enables new business-model innovation across all industries. Slicing networks enables the network operator to maximize the use of network resources and service flexibility. Some of the network functions can be shared between network slices, while other network functions are deployed only for a specific service within the slice. Features like these enable CSPs to offer innovative services to enter new markets and expand their business.

The IBM Cloud architecture center offers up many hybrid and multicloud reference architectures, including AI frameworks. Look for the IBM Network Automation reference architecture and the IBM Edge Computing reference architecture.

Special thanks to Sanil Nambiar for reviewing the article.

Please make sure to check out all the installments in this series of blog posts on edge computing:

• Part 1: “Cloud at the edge”
• Part 2: “Rounding out the edges”
• Part 3: “Architecting at the edge”
• Part 4: “DevOps at the edge”
• Part 5: “Policies at the edge”
• Part 6: “Models Deployed at the edge”
• Part 7: “Security at the edge”
• Part 8: “Analytics at the edge”
• Part 9: “5G at the edge”
• Part 10: “Clusters at the edge”
• Part 11: “Automation at the edge”
• Part 12: “Network slicing at the edge”
• Part 13: “Data at the edge”
• Part 14: “Architectural decisions at the edge”
• Part 15: “GitOps at the edge”
• Part 17: “Storage services at the edge”
• Part 18: “Cloud services at the edge”
• Part 19: “Distributed cloud: Empowerment at the edge”
• Part 20: “Data sovereignty at the edge”
• Part 21: “Solutioning at the edge”
• Part 22: “Connected products at the edge”
• Part 23: “Foundational models at the edge”

Learn more

• IBM Garage
• IBM Edge Application Manager
• IBM Cloud Pak for Network Automation
• IBM Cloud for Telecommunication
• 5G and Edge Computing

Related articles

• “What Is Network Slicing?” — Blueplanet
• “Network Slicing” — Ericsson