Cellular IoT Explained: Implementing Cellular Connectivity

Implementing cellular connectivity for IoT applications; what are the options?

Design engineers must weigh up a broad range of technical considerations when creating a cellular IoT product – hardware architecture, data type and volumes, power consumption, and antenna design, to name but a few. This article provides some best practice advice on developing a cellular IoT product.

Connecting the IoT

The IoT and its industrial cousin, the Industrial Internet of Things (IIoT), continue to gain adoption across our society. IoT deployments deliver tangible business value to industrial processes, optimise production, and streamline supply chains. In our homes, smart home appliances and digital assistants help automate routine tasks, keep us informed of upcoming events, and ensure that our home is secure.

Connectivity is a crucial aspect of any IoT device. Many home, workplace and industrial units typically benefit from Wi-Fi connectivity. However, not all IoT devices have a fixed location. Instead, they are deployed in remote places, on moving vehicles, or across large urban environments. For these, connectivity via the cellular network provides the ideal solution.

Cellular connectivity for the IoT - what are the options?

Cellular communication technologies continue to advance, with 5G deployment underway in most cities. Hot on the heels of 5G, standards for the sixth generation of cellular communications, 6G, is in review and undergoing national adoption and approvals verification. The previous generations of cellular technology, 3G and 4G, provide extensive levels of coverage in most parts of the world, offering asymmetrical download data rates of up to 7.2 Mbps (3G) and 150 Mbps (4G). By comparison, 5G download data rates up to 1 Gbps are possible already, with 10 Gbps likely in the future. Speed is not the only consideration, however. Low latency is essential for some IoT applications, particularly those involved in real-time control, such as industrial robots. Typically, 3G has a latency approaching 1 second, and 4G offers a round-trip latency of up to 200 milliseconds. 5G, already identified as an ideal candidate for industrial control applications, promises sub-one millisecond latency.

Depending on the use case, 3G, 4G and 5G offer exceptional data rates for IoT applications. However, such high bandwidth capabilities are more suited to continuously streaming media files or sending large data files. Most IoT devices, such as environmental sensors or smart city lighting controls, only need a fraction of the available bandwidth.

The asymmetrical narrowband LTE-M (Cat-M1) and NB-IoT (Cat-NB1) cellular protocols have been developed specifically for low-power, low-bandwidth IoT applications. NB-IoT offers 60 kbps uplink and 30 kbps download and a latency characteristic of up to ten seconds. By comparison, LTE-M delivers better throughput of 375 kbps uplink and 300 kbps download and sub-ten millisecond latency, making it suitable for real-time applications.

Architecting cellular-based IoT connectivity

The choice of which cellular IoT connectivity method to use depends heavily on the application. The engineering team will face several technical decisions to implement the selected protocol, including build vs. buy, required range, power consumption, and security.

Build vs. buy: There are two ways to provide cellular connectivity to an IoT device: build a cellular wireless transceiver from scratch or buy a pre-approved wireless module. Designing a cellular transceiver from scratch might appear to offer the most flexibility; however, it is fraught with challenges. Creating any radio frequency circuit function requires extensive experience and is considered a specialist skill. Also, the cellular spectrum is highly regulated, requiring regulatory type approval and network conformance testing according to cellular industry standards stipulated by 3GPP. So, while the reason for building a cellular transceiver from scratch may be well founded, the time taken to perfect a design, put it through type approvals and receive conformance certification would be excessive. A more prudent design approach is to buy a pre-certified cellular wireless module.

Cellular modules offer a low-cost, quick, and flexible method of provisioning cellular connectivity. Most narrowband cellular IoT modules are highly integrated, incorporating all wireless transceiver functions, RF front-end, power management and SIM functionality. The IoT device host microcontroller connects to the transceiver's processor via serial bus interfaces such as SPI or I2C. Also, as traditional SIM (subscriber identity module) cards make way for eSIMs (embedded SIM) and iSIMs (integrated SIM), the advantages of a card-free approach are substantial. Rather than having the burden of managing and deploying hundreds of SIM cards, eSIM/iSIM-based wireless modules significantly simplify the task.

Antenna design: Some cellular modules integrate one or more antennas; however, the diversity of IoT use cases might dictate an antenna external to the module or the device enclosure to achieve the best performance. Highly integrated system-on-chip (SoC) microcontrollers, which include the wireless transceiver, the device host processor, and the antenna in a single substrate (antenna in package), are becoming increasingly popular. Like any wireless application, placing the antenna free of obstacles and away from buildings and vegetation offers the most reliable wireless link performance. Determining the need for an external antenna requires analysing the device's likely geographical coverage requirements, and wireless module specifications (transmit output power and receiver sensitivity) and calculating the worst-case link budget.

Power consumption profiling: Most cellular-connected IoT devices are battery-powered. Estimating battery life is an essential aspect of the design process. Even though NB-IoT and LTE-M are ultra-low-power wireless protocols, predicting how long the device can operate before replacing or recharging the battery requires knowledge of several factors. Power peaks typically occur when the IoT device is processing and sending data. Therefore, its duty cycle - the ratio of time spent active versus time in sleep mode - will determine the average current consumption and, thus, the likely battery life. Developers can avoid power peaks by carefully scheduling embedded tasks and fully using sensor, microcontroller, and wireless module sleep modes. With the dynamic nature of IoT device operation, power consumption monitoring requires specialist power profiling tools. Traditional digital multimeters are unsuitable for this purpose and typically do not have the dynamic range and resolution required.

Security: One of the benefits of using cellular-based connectivity for an IoT device compared to unlicensed wireless protocols is the high level of encryption available throughout the network. Security continues to be a crucial factor for any IoT application, so increasing levels of encryption and authentication algorithms are paramount. Highly integrated wireless SoCs and modules typically incorporate security functions and assist in achieving robust and resilient end-to-end protection from adversaries.

Cellular IoT connectivity - the perfect communications protocol for the IoT

Low bandwidth, low-power cellular IoT protocols like NB-IoT and LTE-M are suitable for many IoT applications. Rather than attempt to architect a cellular transceiver from scratch, specifying an off-the-shelf, pre-approved cellular module will simplify the design's architecture and speed up the development timetable.

Articles in this series:

• Cellular IoT Explained: A Quickfire Guide to Mobile Connectivity
• Cellular IoT Explained: The Pros and Cons
• Cellular IoT Explained: Implementing Cellular Connectivity
• Cellular IoT Explained: Five Ways Cellular IoT is Transforming Our Lives