carloscastilla - Fotolia
What does 5G IoT technology mean for future enterprises?
The emergence of 5G guarantees new architectures to connect billions of IoT devices but also introduces a host of new development challenges for practical applications.
5G is an emerging set of cellular technologies, specifications and proposed standards that promise to dramatically improve the speed and responsiveness of wireless networks. One area that will greatly benefit is the internet of things. Touted benefits of 5G for IoT include faster connectivity, lower latency, reduced power consumption, increased reliability and better security models. Other promises of 5G include improving support for new architectures like IoT edge computing and public mesh networks.
Enterprises are looking for ubiquitous connectivity and near-real-time remote diagnostics and management capabilities for mission-critical IoT -- and 5G is here to answer that call.
Latency and capacity restrictions as well as the use of many proprietary technologies have limited the ability to securely deliver near-real-time industrial services. But 5G could open the doors for a range of new services in manufacturing, healthcare, transportation and public safety.
Ian Hood, chief technologist at Red Hat, said "The implications of 5G will be felt across all industries at the network edge, with factories and warehouses using the industrial internet of things [IIoT] and digitalization to become much more agile and efficient." 5G will allow smart factories to apply advanced visualization, rapidly changeover production lines and increase efficiency and employee safety.
How do 5G and IoT converge?
Developers will be able to access different kinds of network capabilities provided through a unified IoT 5G network, said Colin Alexander, director of wireless infrastructure at Arm Limited. Use cases like enhanced mobile broadband will target more efficient broadband access to connected homes and mobile devices, whereas massive machine connectivity use cases will support the connectivity of billions of small, non-real-time IoT sensors for a range of connected applications. Another class of uses will cover ultra-reliable, low-latency connectivity and target high-value, industrial IoT and vehicle-to-X connectivity that requires a mix of low latency and high reliability.
Don't get distracted by IoT and 5G's promises
In the long run, 5G infrastructure could unify competing wireless standards, said Bruce Collins, director of product management at Cambium Networks Ltd., a wireless infrastructure provider. "It is also adding some confusion in the market today as there are existing private and public platforms being deployed now while 5G is a couple years away from adoption and initial deployments," he added.
5G represents a further splintering or fragmenting of the IoT wireless space, Collins said. Many existing low-power wide-area technologies -- such as LoRa, Cat M1 and Narrowband IoT -- have a track record of successful deployment, while 5G is still new.
Others believe, however, that 5G could eventually replace these older technologies. Danny Tseng, director of technical marketing at Qualcomm, holds that 5G-including variants, such as Narrowband-IoT and enhanced machine type communication, provide many benefits over older radio standards like LoRa and LTE Cat-1. 5G provides better scale, flexibility and security compared to LoRa and can also cover more use cases with more flexibility, efficiency and capability than LTE Cat 1. For example, 5G can scale all the way up to multi-Gbps use cases like ultra-high definition security cameras down to very low-complexity ones like temperature sensors, while Cat 1 can address just a small portion of the IoT need.
The key thing for operators to keep in mind as the options proliferate is that the use is more important than the network technology. Enterprises should focus on identifying a manageable, scalable, secure network that uses the right technique for the right system. "We recommend that enterprises looking to deploy IoT networks start with the application first and then work backwards to the selection of a wireless technology," Collins said.
New architectures for IoT and 5G
5G technologies will provide a variety of options at the transport layer. The use of software-defined networking tools could help future-proof code built for the sensor and application layers.
Existing mesh networks rely on disparate networks to connect to the internet via proprietary gateways to reduce cost and power requirements, but this also adds complexity. If 5G lives up to its promise, enterprise architects might consider replacing these mesh networks with less complex connections direct to cell phone towers, or by using microcell towers installed indoors. "This should increase bandwidth and resilience," said Saar Yoskovitz, CEO of Augury, an industrial equipment monitoring provider.
Tseng said multihop mesh network capabilities are planned for future 5G specifications that will extend IoT and 5G networks beyond the limits of public infrastructure. This could allow a device that is out of the traditional coverage of a 5G network -- for example, in a basement -- to connect to the network using other 5G devices as a relay. 5G new radio IIoT has been proposed as a new standard for industrial apps that could start to roll out in 2021, Tseng said. This could help 5G become the default choice for industrial networking. "The biggest challenge is also having the entire IIoT industry working together to make this happen, but it is making good progress in 3GPP, the standards body responsible for designing the technologies that drive global cellular standards," Tseng said.
5G Release 16 occurred in July 2020, and Tseng expects to see deployments to start in the 2021-2022 timeframe. It includes features for enhanced reliability to support demanding industrial use cases and also includes specification for 5G NR C-V2X for automotive uses. The industry is beginning to look to 5G Release 17, which includes new specifications like NR-Light, expanding the reach of 5G to new types of devices.
'Cloudifying' the network to embrace 5G for IoT
Current mobile technologies are optimized for downloads. But in most IoT applications, data is generated at the edge of the network and uploaded to the network. Said Ouissal, CEO of Zededa, a distributed cloud infrastructure provider for IoT, said he believes 5G network providers will need to invest more resources to create wide-area mesh networks that allow data to flow between devices for safety and situational awareness applications. The opportunity is to design data traffic routing that is more efficient for edge-to-edge communications.
Today, the communication from point A to point B still goes up to the service provider core network and then back down, even if point A and point B are 10 feet apart. "This doesn't work for IoT and strikes at the heart of the new networks' use cases to come," Ouissal said. Data infrastructure providers will need to create networking infrastructure that works more like the cloud than traditional telco networks. If they are successful, this would reduce the burden for IoT developers to think about scaling and provisioning across public networks in the same way that traditional cloud infrastructure reduces concerns about compute and storage infrastructure.
Ouissal said two of the biggest 5G growth drivers have been the further definition and implementation of private 5G networks and the availability of new spectrum in the 3.5 GHz band. "One of the biggest challenges for private wireless networks has been the lack of a well-regulated spectrum, which is now resolved," he said.
The ability to create mini-5G private networks -- including the control plane and data plane that can run locally on premises -- over more reliable spectrums has cleared the path for private 5G to become a viable wireless networking option in operational technology and industrial environments. This, in turn, enables better edge-to-edge communications and a better cost-performance ratio versus traditional wireless technologies like Wi-Fi, Ouissal said.
He said he believes that improvements in edge computing architecture which can preprocess data locally will accelerate the adoption of 5G for IoT applications, resulting in a massive data reduction. For example, an edge node can analyze a video feed locally and only send the recognized objects to the cloud. "5G, edge computing, AI and cloud together are solving real customer problems and enabling customers to achieve their aggressive and important digital transformation goals," stated Ouissal.
