Wireless support is now a necessity for many embedded applications as customers increasingly look to access the benefits of the Internet of Things (IoT). Wireless connectivity provides the easiest path to the cloud allowing access to the IoT for a broad range of applications. Through wireless support, there is no need to run cables to the location of each device and smart protocols such as Bluetooth make it easy to set up devices from practically anywhere.
Wireless Technology Options
Engineering teams now have access to a wide range of wireless connectivity options appropriate for their application.
Short-range: For short-range use in the home or office, Bluetooth and WiFi span the range of features required for low power and high bandwidth. Even in larger installations, such as within a large warehouse or shop floor environment, the use of WiFi gateways and mesh networking in the case of low-power networks such as Bluetooth or Zigbee, greatly extend the operational range.
Mid-range: Wireless networks such as LoRaWAN and Sigfox make it possible to access devices at ranges of tens of kilometres at a very low cost. These networks are designed to transport small amounts of data, supporting data of rates up to 50kb/s in the case of LoRaWAN, and unlike cellular services, these networks are not billed by the byte. Users can operate their own gateways, which may be required for service in remote locations, but also provide options where cellular coverage may not be reliable. Service providers such as The Things Network have built public networks based on LoRaWAN that span entire cities and their suburbs. The Things Network now has more than 13,000 gateways in operation globally including India with 58 gateways at the moment. The number is bound to grow in future as India explores endless possibilities with LoRaWAN.
Long-range/widespread: When more widespread coverage is required, and the application can support pay-per-use billing, designers can make use of 2G, 3G and 4G cellular networks.
The history of wireless connectivity
A problem faced by many designers in the past was that the development of efficient RF interfaces to enable wireless connectivity is difficult and time-consuming. The high-frequency signals that need to be passed from the antenna to the transceiver need delicate handling to ensure a high signal-to-noise ratio. The antenna itself can have a dramatic impact on performance, with small changes in the shape and structure meaning the difference between an RF transceiver that works at the desired range and one that is only effective over short distances. In addition, wireless systems need to pass stringent tests that determine whether the system will interfere with other users, even in bands that do not need a specific radio licence, such as the 2.4GHz bands employed by Bluetooth, WiFi and others.
Traditional single-board computers (SBCs) did not have wireless connectivity, forcing design teams to develop custom modules but as ecosystems around Raspberry Pi, BeagleBone and Arduino developed, ready-made modules that could be used with the core SBCs were developed. Options included, for example, the WiPi module for the Raspberry Pi, connecting to the processor module through a USB port. The BeagleBone and Arduino made provision for similar wireless module options, connecting through pin-header connectors.
Although the use of add-on wireless modules reduced the time associated with hardware design, it did not cut out the time associated with other aspects of wireless integration. The designer would, in many cases, have to pursue their own testing to ensure compliance with legislation covering RF emissions in the locales where they wanted to sell products.
A further demand in terms of time and cost came from the software integration process. There are many choices of transceiver silicon, especially with popular short-range, low-power wireless protocols such as Bluetooth or WiFi. The protocol support on the wireless module is provided by a dedicated SoC that runs a number of firmware libraries itself, but in some cases, high-level protocol layers may have to run on the host SBC, with some form of proprietary serial protocol used to pass data between the layers. The integrator would be responsible for bringing these software elements together in a cohesive manner.
Today’s solutions for wireless connectivity
As wireless connectivity has become more popular, SBC manufacturers and integrated hardware-software development environments such as Arduino have taken a variety of paths to bringing wireless support to their product lines. The choices include integrated solutions that take advantage of the integration provided by a growing range of microcontrollers that incorporate direct support for wireless protocols. Others employ modular architectures to provide designers with a choice of wireless connectivity options that can be used with a common base board.
Using integrated module bring several key advantages not least of which is that the complete SBC, including wireless connectivity, has been tested for compliance with RF emissions legislation in all the territories for which the modules are available for sale. The engineering team can also take advantage of pre-integrated firmware that can, in some cases, make wireless connectivity as simple as sending data through a serial port and often, the integrated solutions offer lower energy consumption than combinations of base boards and addon modules because the suppliers have been able to take full advantage of the processing that is available on both the core MCU and the processor cores inside the wireless transceivers.
The result of the integration is lower overall development time and smaller form factors compared to designs based on sandwiches of multiple boards. Very often the cost is reduced due to the higher integration and developers can continue to take advantage of header-pin and similar interfaces to attach custom sensor modules.
Developers have access to powerful high-integration solutions such as the Raspberry Pi 4 Model B Computer. It offers ground-breaking increases in processor speed, multimedia performance, memory and connectivity (in comparison to the prior-generation Raspberry Pi 3 Model B+), while retaining backwards compatibility and similar power consumption. The product’s key features include a high-performance 64-bit quad-core Cortex-A72 (ARM v8) processor, dual-display support at resolutions up to 4K via a pair of micro-HDMI ports, hardware video decode at up to 4Kp60, 4GB of RAM, dual-band 2.4/5.0GHz wireless LAN, Bluetooth 5.0, Gigabit Ethernet, USB 3.0, and PoE capability (via a separate PoE HAT add-on). The dual-band wireless LAN and Bluetooth have modular compliance certification, allowing the board to be designed into end products with significantly reduced compliance testing, improving both cost and time to market.
The BeagleBone Black replaces the Ethernet controller of the original SBC design with 2.4GHz WiFi interface and a Bluetooth 4.1 and BLE transceiver. The integrated module is built around Octavo Systems’ SIP. This couples an Arm Cortex-A8 processor with 512GB of high-speed DDR memory and 4GB of flash. To support applications that need high-speed I/O processing, the TI-made processor is supported by two programmable real-time units (PRUs). The PRUs offload tasks that need low-latency processing from the Arm processor, providing greater headroom for an operating system, user interface and system-management functions.
For simpler designs, the Particle Photon couples a Cortex-M3 microcontroller from STMicroelectronics with a Cypress WiFi controller: the same type as that used in smart-home devices such as the Nest Protect and Amazon Dash. The Particle Electron takes the same core processor complex and applies it to a 3G cellular transceiver, providing the ability to build IoT nodes that do not need a local gateway to connect to the cloud.
Modular solutions provide another route to bringing wireless connectivity to an SBC-based system. With Arduino’s range of products, the development team can choose from a variety of modules – known as Shields in the Arduino ecosystem – to add an RF interface to a base board. The Shields in the MKR family add local or wide-area wireless network connectivity. The MKR 1000 and 1010 both include a WiFi transceiver. The WAN 1300 provides LoRA connectivity and the GSM 1400 access to the many cellular networks available around the world. In addition, the MKRFOX 1200 acts as an interface to the SigFox low-power wide-area network.
Each of these Shields can be mounted on a carrier board such as the Genuino Zero or the Due using the pin-header connectors. A notable feature of many of the MKR family of modules for wireless connectivity have their own 32bit microcontroller based on the Arm Cortex-M0+ core. Developers can take advantage of the M0+ to offload packet processing, such as encryption and compression, from the carrier’s main processor. Alternatively, devices such as the MKR1000 can be used as standalone modules in space-constrained systems, with the pin headers removed to reduce overall volume.
Development kits provide other options for engineering teams, mainly for those who want to design custom modules for volume production. Development kits are designed to get teams up and running as quickly as possible. For example, the element14 development kit for the TI SimpleLink Sensor to Cloud Linux Gateway provides an end-to-end solution. The kit contains all the components needed to create a full sensor network, including a gateway solution based on the BeagleBone Black board augmented with WiFi as well as a CC1350 LaunchPad kit to act as a long-range sensor node. Other options for development from TI include the CC3200, with support for WiFi, and the CC2650, which targets Bluetooth, ZigBee and 6LoWPAN networks. Thanks to a rich portfolio of platforms that range from prototyping kits to off-the-shelf SBCs that can be used in production systems, engineering teams can now take full advantage of the novel business models made possible by wireless connectivity in the age of the IoT and deliver highly differentiated solutions without having to deal with the complexities of RF design.