5G networks must support a variety of very diverse use cases with different requirements for latency, throughput, and availability. The dynamic network slicing concept offers a way to optimize 5G networks to address all use cases efficiently.
Dynamic network slicing enables the design, deployment, customization, and optimization of different network slices running on a common network infrastructure. It leverages innovations in cloud mobile access and core. And it capitalizes on the capabilities of software-defined networking (SDN), network function virtualization (NFV), end-to-end orchestration, network applications, and analytics.
5G must enable and support new services
Today’s networks were designed for the delivery of personal communication services and content, such as voice, video, and web browsing to a variety of devices. But 5G networks must enable and support a fully mobile and connected society.
With 5G, a proliferation of wearable devices and wireless connected objects will pave the way to a wide variety of new uses. More pervasive human-centric applications will be launched, such as virtual reality augmentation and UHD video. At the same time, machine-to-machine (M2M) and human-machine type applications will be introduced that will make our lives safer and smarter. All these applications will automate and mobilize a variety of industries, including energy, health, public safety, smart city, manufacturing, logistics, media, and automotive.
Beyond connectivity, 5G will give operators unique opportunities to create new business models for consumers, enterprises, and industry specific services, as well as content and application providers.
New end-to-end mobile network architectures needed
As noted by the NGMN, the success of 5G technology will be based on the ability of operators to provide multiple solutions for all requirements. And the ability to provide all stakeholders with a unique solution tailored to their specific needs.
Unfortunately, today’s network architectures are not up to the challenge. The current “one-size-fits-all” approach to wireless networks for all use cases and services to every device everywhere is no longer viable. It does not offer the level of adaptability needed to meet the performance expectations for new and legacy use cases, services, business models, infrastructure usage approaches, and radio access needs that will emerge with 5G.
To support new communications demands from a wide variety of users, machines, industries, and other organizations, mobile network architectures must evolve. They must move from the current network of entities architecture to a network of capabilities architecture. And they must enable a shift from the current network of connectivity model to a network for services model. The key to this shift lies in how end-to-end 5G networks will be designed, architected, implemented and operated.
Dynamic Network slicing enables end-to-end performance
Subscribers measure service satisfaction based on end-end performance, they are not aware or concerned if an application is running well in the radio, transport, core or the cloud.
To meet future performance expectations, the individual network segments (radio access network (RAN), transport, metro, core, edge cloud, central cloud), which were formerly treated completely separately, must be re-evaluated. Their performance must be adapted and coordinated to deliver a specific flow, for a specific user, at specific time.
Dynamic network slicing offers an effective way to do that, meet all use case requirements, and exploit the benefits of a common network infrastructure. It enables operators to establish different deployments, architectural flavors, and performance levels for each use case or service group and run all network implementations in parallel.
Dynamic End-to-end network slicing realizes a fully programmable network.
With dynamic network slicing, operators could create a fully programmable network architecture which suits the requirements of various use cases, subscriber types and apps. They could allocate dedicated and optimized end-to-end virtual network functions and physical resources for each use case or group. And within a slice, the functional elements could be instantiated according to the needs of a specific service offering.
This is achieved by analyzing service requirements and identifying KPIs. With this information, the right network resources could be allocated at the right time and in the right place based on performance requirements, the QoE needed, and the relevant service quality indicators that must be monitored and assured.
For example, latency-critical services could be allocated functions in the radio access or edge cloud to minimize delay. But to improve efficiency, these same functions could also be allocated in the central cloud for other services.
NFV and SDN enable network slicing
The dynamic network slicing concept leverages NFV and SDN to create many dedicated end-to-end virtual networks. All end-to-end network slices are created and operated over a common physical infrastructure.
A slice is self-contained. It has all the functions and capabilities, appropriately chained together to best meet all the needs of the corresponding services and use-cases.
Dynamic network slicing could also be used to support a variety of business cases. For example, ‘multi-tenancy’ in which mobile network tenants could share the same network infrastructure with vertical industry tenants. Each tenant could operate and manage their corresponding network slice by exploiting built-in network slice customization and optimization capabilities.
With this approach, the network could be sliced:
- Per type of customer services when an operator wants to optimize dedicated services for end users (e.g., tactile internet or ultra HD video) or wants to address industry-specific customer needs (e.g., e-health, sensor network, high speed train)
- Per tenant when an operator wants to share network costs (e.g., network platform as a service, network infrastructure as a service)
Managing and orchestrating dynamic end-to-end network slices
An end2end network slice orchestrator is envisioned to manage all aspects of network slicing. Such an orchestrator will help to manage the entire life cycle management of a slice, from design and creation to operation and optimization.
Slice management and orchestration would be applied at different levels — across slices and within the slice — and at different times. And network resources would be allocated to a slice based on a tradeoff between guaranteeing resources to an individual slice and the advantage of pooling resources from all slices.
Finally, automated processes would be used to manage the entire life cycle of all slices:
- Monitoring of network functions, services, and loads on virtual and physical nodes with well identified and standardized KPIs
- Proactive slicing of policies to support decision matching to customer/tenant expectations based on network analytics
- Configuration and reconfiguration using inputs from analytics and from monitoring specific performance requirements established by policies
Dynamic end-to-end network slicing optimizes the entire 5G network
Dynamic end-to-end network slicing provides an optimal approach to network slicing in 5G networks. It partitions the same network infrastructure into multiple logical end-to-end network slices. With these slices, operators could support the very diverse and extreme requirements for latency, throughput, and availability needed to deliver 5G services to a wide variety of users, machines, industries, and other organizations.
With this innovative approach, operators could create different deployments and architectural flavors for each use case or service group. They could run all network implementations in parallel, simultaneously on a common network infrastructure. And they could open the door to new communications possibilities and use cases, many of which are as yet unknown.
White Paper: Dynamic End-to-End Network Slicing for 5G
Nokia 5G Masterplan