The Internet of Things (IoT) phenomenon—ubiquitous connected things providing key physical data and further processing of that data in the cloud to deliver business insights— presents a huge opportunity for many players in electronics and software, including chipset vendors, device developers, OEMs, manufacturers, equipment vendors, network operators and end-to-end solutions providers. Many companies are organizing themselves to focus on IoT and the connectivity of their future products.
For the IoT industry to thrive, three items are crucial: a viable business model, a robust connectivity topology and reliable devices. This article discusses these, focusing on the design challenges that must be overcome to make reliable devices. Challenges vary depending on the IoT application. While cost is a major factor in consumer applications (e.g., wearables and home automation), industrial IoT applications (e.g., smart grids, connected cars and transportation) require unfailing reliability, longevity, security and the ability to operate devices with little or no human intervention.
The Business Model
End-to-end solution providers operating in vertical industries and delivering services using cloud analytics will be the most successful at monetizing a large portion of the value in IoT. Low power, wide area (LPWA) IoT technologies open up possibilities for service providers. Knowing the location of pets and vehicles, tracking valuable personal belongings, monitoring utility usage, obtaining real-time data on the health of crops and livestock, employee fatigue and machine status are useful for individuals and businesses.
A typical smartphone contract delivers roughly five cents per MB of data. Assuming an IoT application uses 100 KB per month, and a user is willing to pay a modest 10 cents per month for these new IoT applications, that’s already better business for an operator. Delivering $1 per MB is 20x more revenue than a typical smartphone contract for the same amount of data consumption. While many IoT applications may attract modest revenue, some can attract more than $10 per month. For little burden on the existing communication infrastructure, operators have the potential to open up a significant source of new revenue using LPWA technologies. Clearly, it is important to understand the value chain and business model for the IoT application.
Figure 1 shows a simple IoT network model, consisting of a device layer containing “things” with sensors and actuators that capture or initiate physical events. These connect to gateway devices using short-range wireless links, and the gateways communicate to the cloud via wide-area networks, such as LTE.
Across a variety of vertical industries, the realization of IoT networks will involve a heterogeneous mix of wireless technologies, including NB-IoT, Cat-M, Z-Wave, ZigBee, SIGFOX, LoRa, ANT, Thread, Wi-SUN, Bluetooth and Wi-Fi. Individually and collectively, these pose special challenges such as power dissipation, transmission range, data rates, seamless connectivity, handshake protocols, security and radio compliance. This diversity in deployed technologies presents a significant opportunity and challenge for the entire IoT industry. Modern microcontrollers make it possible for machine learning to run on even the lowest power devices at the edge of the network, to respond to sensor data and send triggers when actionable events take place. Connectivity topology becomes more interesting with distributed machine learning, analytics and intelligence in gateways and end nodes making more efficient use of bandwidth.
Reliable Devices and Design Challenges
IoT devices present many design challenges, some similar and many different than with smartphones. Developers must overcome constraints from battery drain, power, signal integrity and the complexities of the RF chain. LPWA technologies such as NB-IoT are governed by 3GPP, which requires RF conformance testing before being deployed on networks. Interference and coexistence must also be verified.
The following sections dive deeper into each of these design challenges to producing reliable devices (see Figure 2).
Optimizing and guaranteeing power consumption is a requirement for many IoT devices. In some installations, multiple years of battery life may be committed through a service level agreement (SLA) contract. A software update could use months of battery capacity, and too many “over the air” updates to resolve defects and security issues could compromise battery life. Network settings and handshake protocols between the device and the network can also reduce battery life significantly. What happens if the network is down? Does the device search repeatedly for the network and drain the battery?
For IoT devices, the active state —when the device is transmitting or receiving data—is very short compared to idle and standby states. Measuring the current consumption is key to understanding and optimizing the power consumption. Figure 3 illustrates the different device operating states and the resulting current drain, likely a ratio of 1:1,000,000, from sub-µA to 100 mA. For example, in transmit, Bluetooth low energy transmitters use tens of mA compared to a few A for GSM transmitters. The majority of the time, devices are in idle mode, drawing from tens to hundreds of nA up to hundreds of µA.
Because IoT devices have very low duty cycles, a common way to lower the total current drain is to design the system so the device has a very short active state followed by periods of relatively low activity or no activity. The challenge in verifying the likely battery life is to accurately measure the dynamic current drain across the different operating modes over a period of time and with a single view that provides a complete and detailed analysis.
With many types of devices deployed in consumer and industrial applications (e.g., smart grid, smart energy, smart factories and smart homes), many IoT formats are being deployed and many operate in the same spectrum (e.g., Wi-Fi, Bluetooth and ZigBee). These environments will affect multi-radio interference (co-channel or adjacent channel), transmission range and speed, and interoperability. All must be considered.
NB-IoT, Cat-M and other LPWA technologies use narrow bandwidths to connect to IoT/M2M devices, resulting in lower data rates and low power. Thread is a secure wireless mesh network for home and connected products; based on the 802.15.4 physical and MAC layers it enables the gateway to easily control connection to the cloud. Bluetooth has been the most commonly used format for consumer electronics and is often used around the smartphone and near field communication (NFC) for payments. Wi-Fi is a foundational technology that is used wherever possible. LPWA formats like LoRa and SIGFOX are opening up new applications. Many other technologies such as ZigBee, ANT and Z-Wave are being used for home automation.
With all these technologies, the radio design has to be optimized for data rates and sensitivity. Good RF transceiver and antenna design are needed to achieve deep in-building coverage, as making a receiver work hard to decode a weak signal further reduces battery life.
Power and Signal Integrity
The increased demand for expanded functionality in a small form factor drives the need for higher density, lower power and compact circuit design. Maintaining signal integrity and power integrity becomes more challenging as traces get closer together and supply voltages are lowered.
Common signal integrity issues that can degrade overall system performance include reflections, excessive losses, crosstalk, distortion and power supply noise. Power integrity speaks to how effectively the power is converted and delivered from the power source to the load within the device. With the drive toward low power electronics, DC supply voltages and tolerances have been reduced, some from ±5 percent to ±1 percent. Ripple, noise and transients riding on the output power rails can adversely impact the clock and accuracy of digital data. Designers need a power integrity solution that can measure these low DC voltages with high accuracy.
Co-Existence and Conformance
IoT devices may be integrated into products and gateways that include a variety of different standards operating over a range of frequencies. Each radio and device type will need to meet a specific set of downstream acceptance tests. Cellular devices need to pass certification tests from standards bodies, and many operators have their own acceptance test plans. All devices must pass regulatory testing that depends on frequency band and region. Many system integrators run their own acceptance tests to select modules in their systems. So designers need to ensure that interference and intermodulation effects are anticipated, understood and tested.
IoT devices at the edge of the network increase the security vulnerability of networks. Yet devices don’t have the needed resources to host sophisticated security protocols. The capability for remote software updates needs to be designed into the device to allow for security updates, which has implications on battery life. Authentication and cyber security features need to be tested and upgradeable. The most secure development approaches available today are likely to be compromised in the coming years, so security patches should be anticipated. In the future, security will likely be built into these devices.
IoT is likely to be a significant enabler of many disruptive business models and market efficiencies. Recall how the internet and players like Amazon, eBay and Uber have and are transforming markets. Peer-to-peer banking, personalized car insurance, personalized health insurance and crowd-sourced businesses are emerging. IoT devices for ordering convenience will lead to more services on top of products. These new business models and services rely on networks of sensors and actuators, linked by radio and connected to the cloud for data analytics. For these big connected sensor systems to work well, even the smallest of components must be secure, stable and reliable.
Kailash Narayanan is vice president and general manager at Keysight Technologies, where he leads the wireless device business and is responsible for product and solution development, marketing and delivery to the wireless device ecosystem. His current focus is 5G and IoT. He received a master’s in electrical engineering from the University of Illinois, Chicago and an MBA from Walden University.