The emerging 5G network offers exponentially more capacity, lower latency, as well as increased reliability and availability. As such, 5G will play a key role in the growing IoT ecosystem, supporting a wide range of IoT verticals, including smart cities, automotive, health care, transportation, and industrial automation. These verticals will support a wide range of applications and use cases – each with a diverse set of performance and service requirements. Yet these varying requirements present service providers with new challenges, requiring the underlying network infrastructure to meet a variety of different functionality and performance requirements.
A key element in making all this happen is the ability of 5G to adapt network resources in both the user plane and the control plane to meet the performance levels of each service and device. Network slicing has been devised as a way to structure a network to support diverse classes of services in a guaranteed way on the same network.
By “slicing” the single physical network into multiple, virtual, end-to-end networks, a service provider can software-define the necessary performance criteria – such as speed, security, coverage area, etc. to match the requirements needed by a service using the slice.
With network slicing, network functions and resources that are specifically tailored to a vertical market’s requirements are logically isolated, creating multiple virtual networks on top of a common shared physical infrastructure.
At present, network slices are categorized into different types according to the abstraction of characteristics of the services they facilitate. The Third Generation Partnership Project (3GPP) has defined four network slice types; (1) Enhanced Mobile Broadband (eMBB); (2) Ultra-Reliable Low-Latency Communications (URLLC); (3) massive Internet of Things (mIoT); and (4) V2X.
These slice types are based on the typical characteristics required for use cases and verticals to provide a way for establishing global interoperability for slicing.
Are more “slices” needed?
ATIS recently released an innovative new report, IoT Categorization: Exploring the Need for Standardizing Additional Network Slices, that examines the topic of network slicing. Its objective is to identify areas of commonality across IoT devices and applications and examine how these may be grouped into standard network slices to keep service quality consistent across different networks and operators.
The report considers the network requirements across the multidimensional landscape of IoT devices and applications to identify if any additional network slice types may be defined to ensure consistent service quality across operators.
While existing IoT initiatives take an application-centric approach, often from the perspective of a single application or industry vertical, ATIS’ IoT Categorization Initiative examines IoT from a network-centric perspective to determine an effective way to categorize IoT into a small number of meaningful categories based on device types, applications, services, or a combination of these; as well as business, technology, and regulatory implications.
To provide structure for the analysis, the major IoT application groups are identified and decomposed into further use cases along with the IoT devices used. Key performance characteristics that influence network slice requirements are identified. Each application/IoT device grouping is mapped to a network slice, highlighting any areas that may warrant a new slice type. The resulting “living” IoT Categorization Matrix forms the basis for identifying areas of commonality that can be used to determine if additional network slice types are needed to support the varied IoT applications and devices that will connect to the network.
Standardized definitions for the most commonly used slice types provide a mechanism for enabling roaming, as well as global interoperability for network slicing across network operators.
Through IoT Categorization, a smaller subset of the applications and devices were found to exhibit characteristics across multiple slices that were not a perfect match to the currently defined four standardized slice types. This has resulted in the recommendation for consideration of an additional standardized slice type to address the performance requirements for this subset, which encompasses use cases across industrial automation, robotic surgery, and public safety.
The proposed new network slice type, High-Performance Machine-Type Communications (HMTC), would be defined by the key performance characteristics common to the applications in the above subset, including low latency, high availability, and high data rates. Although similar to the characteristics of the newly defined V2X slice, no mobility or sidelink is required.
5G is envisioned to support a wide range of IoT verticals with a diverse set of performance and service requirements. Network slicing is a key component of the 5G architecture, enabling network operators to expand their customer base and offer new, differentiated services. Through network slicing, operators can allocate their network resources based on a precise set of performance requirements. Defining a sufficient number of standardized network slice types is essential as customers will expect the same level of service no matter where they are. With the addition of the recommended HMTC slice definition, the resulting standardized slice types should address the most commonly used services and associated performance characteristics. Providing this broader set of standard slice types ensures roaming support, as well as global interoperability for network slicing across network operators. In sum, it is part of what it will take to help 5G reach its potential.