Many of the business cases for investment in 5G revolve around the Internet of Things (IoT), both in industrial and residential contexts. Mobile operators, private network providers and enterprises are all assessing the potential of 5G to support reliable, secure and responsive IoT use cases, which could accelerate digital transformation programs or support new customer experiences and revenues.
However, the variety of use cases and traffic types that need to be supported is daunting, and in most ways, it is misleading to refer to a single IoT. Instead, this is an umbrella for thousands of potential applications in hundreds of environments. There are only a few elements in common between them, one of which is small cells.
Small cells are the only way to support all the many connectivity requirements of different IoT services cost-effectively. They can be used to fill gaps and achieve true universal coverage, including indoors. They can provide hotspots of capacity where huge numbers of sensors and devices need to be supported in a small area. And their low power levels fit well with the low energy nature of many IoT use cases.
Rolling out connectivity for the IoT is not a simple matter. There are complex architectural and commercial decisions to make, and because the IoT is in its early stages, it will be important for service providers to take a flexible approach which can enable whatever use cases may emerge in future. With these complexities in mind, Small Cell Forum’s TECH Working Group is compiling an important document, entitled ‘5G era IoT use cases and enterprise small cells’, which will help stakeholders decide on their first priorities for addressing this new opportunity.
The paper advocates a platform approach which will maximize the return on investment in IoT and enterprise small cells, and avoid technology or commercial dead ends.
One of the big dilemmas, especially in the industrial IoT, is whether to deploy connectivity that is fully optimized just for a few use cases (an ultra-low power network specifically for sensor monitoring applications for instance); or a more general purpose network which is flexible enough to support a wide range of use cases, both those which are understood now, and those which evolve in future.
A private enterprise may decide that a specialized single-purpose network is worth the investment for a critical process like smart manufacturing, and private operators are an important and growing group of small cell deployers. Such specialized IoT systems would commonly work in privately owned or shared spectrum and would often have their own local packet core and edge node to provide a self-contained, dedicated network.
But for most service providers, it is clear that each individual IoT use case will contribute only incremental extra revenues. The key to ROI will lie in a platform that can aggregate a large number of applications, which will cumulatively deliver significant new income – which could also be topped-up with cross-app services like security or device management.
Small cells are ideal to support a broad platform of this kind, and the TECH paper outlines how they can underpin a huge variety of services and traffic types, and how they need to be integrated with other technologies like edge computing to maximize their capabilities.
This approach is important because IoT services will need a far more flexible approach to provisioning networks than mobile broadband has done in the past. Some applications will require truly ubiquitous coverage, including deep indoor penetration, and even reaching unpopulated areas such as large farms or remote sites like mines. If the applications are business-critical, such as infrastructure monitoring, those links will need to be secure and reliable. Reaching indoors and out to remote areas is most effectively done with small cells, often combined with their own core and edge node to create self-contained networks optimized for particular IoT applications.
Other IoT services will be quite the opposite and require hotspots of very localized connectivity, while some will need to operate at very low power or very high availability. A network that harnesses small cells in different form factors can address all these needs within a unified management system.
In many cases, traffic patterns will be more variable in different times and places than smartphone usage. The more network resources can be targeted flexibly at areas of demand, the more efficient the platform will be. In future, network slicing may be the way to support multiple use cases with different requirements, on a single infrastructure. In the short term, this can be achieved by integrating many physical resources, including different small cells plus edge and core systems, within a flexible platform for control and orchestration.
Figure 1 shows the SCF’s proposed architecture for an IoT architecture which can support high levels of device density as well as variable traffic patterns and universal coverage. This provides a valuable blueprint for all kinds of service provider, many of which are already grappling with the complexities of how to support IoT connectivity most efficiently for enterprise or residential purposes.
As the figure shows, there are many elements involved in supporting a wide range of IoT use cases, and the diversity will only increase as 5G enables even higher levels of device density plus ultra-low latency and critical availability. Combined with supporting technologies like AI and edge computing, operators can look ahead to advanced cases such as mobile robotics and cobotics, or fully immersive VR for industrial or entertainment use. But the platform must also support more immediate use cases which can generate revenues now.
The small cell networks which support all these use cases will be a mixture of indoor and outdoor, enterprise and home, and may be run by an MNO or a neutral host or private operator. As well as 4G and 5G RAN, there will be other key technology enablers to integrate, such as the 4G and 5G core networks, which by supporting separation of control and user planes (CUPS) will enable much of the flexibility that will be so essential to the IoT business case, making it easier to align the right network resources with each traffic type or application. Edge computing, virtualized networks and, in future, network slicing, will all be able to enhance the business case further as they are introduced to the mix, bringing higher degrees of resource efficiency and flexibility to support an even wider range of services.
As service providers contemplate the dizzying array of IoT use cases, and the uncertainties, at this early stage, about business models and architecture choices, the SCF paper will provide an invaluable starting point, and a practical, business-focused blueprint to help members start on their IoT journey.