A View on 5G NR-IoT devices
_{Reference Design Architecture}

One of the terms considered for 5G today is MAGIC – Massive, Autonomous, Gateway, IoT, and Connectivity. 5G Mobile Communication networks are characterized by a heterogeneous nature, allowing support for features including multiple access technologies, high data speed, and enhanced security (see Figure 1). The 5G system will be based on a complete IP-based network to ensure a 5G NR-IoT device can seamlessly handover and automatically connect to other wireless mobile networks, such as 4G and Wi-Fi, based on user choice and critical IoT application requirements. Handovers in 5G networks are not only carried out to maintain a seamless connection with other active networks, but also to provide better services to end users and to meet individual requirements.

Figure 1: 5G Requirements (Source: Capgemini)

According to market reports, the number of connected IoT devices installed worldwide is expected to exceed 75 billion by 2025. In the near future, the number of new IoT applications and the data they generate will be extremely high due to the increase in customer demands. This will make for a challenging situation, where killer IoT applications will require faster, smarter, simpler, more reliable, and more scalable architecture than before. Hence, an approach is required for designing 5G NR-IoT devices (see Figure 2) to treat current IoT challenges, where security, connectivity, data analytics, data computing, and data management require a different processing model altogether. OEMs, ODMs, and cloud platform providers are already working to address the challenges for building a futuristic end-to-end IoT system.

Figure 2: 5G NR-IoT Hardware Platform Design (Source: Internet)

Some of the new features identified in 5G NR-IoT devices are Edge computing, seamless connectivity, security, simultaneous and multiple service activation for real-time and non-real-time applications, end-to-end Quality of Service (QoS) management under varying network conditions, and unified hardware design for fast switching with multiple access technologies during a network failure. The proposed 5G NR-IoT device software architecture has five new software frameworks and a modular hardware design based on a vision of 5G that includes: –

Figure 3: 5G NR-IoT Device – Reference Design (Source: Capgemini)

• A new connectivity framework to support seamless handover operation, when the user moves across various cellular networks, such as 5G, 4G, and Wi-Fi, with guaranteed QoS based on emerging 3GPP standard ATSSS – Access Traffic Steering, Switching and Splitting
• A new interoperability framework that provides an M4 (Multi-technologies, Multi-device, Multi-protocols, Multi-vertical) type of design model, which enables end users to connect with different commercially available IoT products. 
• A new data analytics framework for monitoring packet transmission mechanisms for killer IoT services, providing reliable QoS and a high data transmission rate according to user preferences. Different services and applications in 5G networks have unique requirements. To meet expectations, dedicated virtual network resources are needed to guarantee service quality for each network slice, where the monitoring mechanism is required to ensure that the required QoS is met.
• A new blockchain framework, wherein the data is stored on the gateway and exchanged in the form of blockchain blocks to support decentralization and ensure the integrity of the data sent and received from and by the network. Since data in the blockchain are exchanged in the decentralized database and encrypted using the TPM (Trusted Platform Module) as a root-of-trust for protecting sensitive home data, a 5G-NR IoT device based on blockchain addresses key security requirements like confidentiality and integrity.
• A new Zero Touch Framework to eliminate the manual provisioning process in a IoT network by configuring the 5G-NR device automatically and by providing seamless connectivity with the ISP / cloud platform provider
• A new modular hardware design based on emerging SoC chipsets covering the application processor, energy harvesting module, PTA (Packet Traffic Arbitration,) and 5G radio transceiver along with other radios like 4G and Wi-Fi as fallback connectivity, enabling 5G IoT devices to be always on.

The proposed 5G NR-IoT device architecture (see figure 3) relies on interactions amongst different layers, spanning the application layer to a unified hardware platform as a plug-and-play framework with open APIs. 5G NR-IoT multi access devices have a larger number of requirements that should be supported, including:

• Operating with minimal inputs from the user. From a user experience perspective, it is preferable that decisions are carried out in an automated way without having to query the user every time a new interface becomes available, or an old interface disappears.
• Selecting interfaces based on user preferences, application requirements, and information about the network.
• Balancing traffic over available interfaces in a way that is transparent to the user, i.e., as seamlessly as possible.

The 5G NR-IoT device reference design provided in this article provides enhanced data security, an ABC (Always Best Connected) model, real time data analytics, and a modular hardware design that can be deployed in any on-premises/business critical applications, thereby creating a major backbone for IoT infrastructure design. The proposed 5G NR-IoT device can be designed specifically to support Industrial IoT (IIoT) and Consumer IoT (CIoT) applications, where the 5G NR-IoT device can deliver ultra-low latency and incredibly high bandwidth connections supporting AI-driven applications by serving the large number of wired/wireless sensors and various instrumentation, equipment, and machines located in the industrial manufacturing plants as shown in figure 4. With distributed architectures created based on 5G NR-IoT devices that can also act as Edge devices, the local data processing allows machine learning algorithms to be applied for massive amounts of data without leaving the security and privacy of the IoT network. The real-time and non-real-time traffic can be managed quite easily at the edge of the IoT network with the help of a 5G NR-IoT device.

Figure 4: Category of applications for building the 5G NR-IoT device (Source: 3G4G blog)

However, one of the key challenges in managing the 5G IoT network is to allocate the network resources for dynamic and critical applications to maximize the benefits of the 5G network characteristics. Network slicing for 5G is generating significant attention in the wireless and telecommunications industries to provide Network-as-a-Service (NaaS) for different use cases. Network service providers can use the network slicing concept to generate new capabilities by defining specific latency, performance, and reliability for different applications such as use-case-based slice instantiation, and SLA-driven slice instantiation (dedicated or shared) either at the device, at the edge, or at the core of the network. The network slices can run in isolation from each other, and each and every slice is used to solve a specific business problem in the IoT ecosystem.

As a summary, the 5G NR-IoT device shall enable

• User-centric ->uninterrupted connectivity and communication services; smooth consumer experience
• Service provider-centric ->connected intelligent systems; multi-tenant, multi-domain capabilities; large area of IoT services; critical monitoring/tracking services
• Network operator-centric -> scalable, energy-efficient, low-cost, efficiently managed, programmable, and secure; communication infrastructure

The 5G NR-IoT device reference design provided in this article provides enhanced data security, an ABC (Always Best Connected) model, real time data analytics, and a modular hardware design that can be deployed in any on-premises/business critical applications, thereby creating a major backbone for IoT infrastructure design. The proposed 5G NR-IoT device can be designed specifically to support Industrial IoT (IIoT) and Consumer IoT (CIoT) applications as shown in figure 4, where the device can deliver ultra-low latency and incredibly high bandwidth connections supporting AI-driven applications by serving the large number of wired/wireless sensors and various instrumentation, equipment, and machines located in the industrial environment. With distributed architectures created based on 5G NR-IoT devices that can also act as Edge devices, the local data processing allows machine learning algorithms to be applied for massive amounts of data without leaving the security and privacy of the IoT network. The real-time and non-real-time traffic can be managed quite easily at the edge of the IoT network with the help of a 5G NR-IoT device.

Vijay Anand

Senior Director, Technology, and Chief IoT Architect, Capgemini Engineering

Vijay plays a strategic leadership role in building connected IoT solutions in many market segments, including consumer and industrial IoT. He has over 25 years of experience and has published 19 research papers, including IEEE award-winning articles. He is currently pursuing a Ph.D. at the Crescent Institute of Science and Technology, India.