Wireless Sensor Networks Will Drive “The Internet Of Things”

Combine billions of IP-enabled (Internet protocol) devices, RFID tags, wireless sensor networks, machine-to-machine (M2M) communications, iPhone apps, white space TV spectrum, and cloud computing, and the result is the practical realization of the Internet of Things (IoT).

Gartner recently added IoT to its “Hype Cycle,” meaning the technology is “on the rise” as a mass-media hot topic. Despite the hype, IoT technology has become in many cases commercially viable, making life easier, safer, and greener for millions of people.

In 2012 and for years to come, IoT will be a significant driver of deeply embedded devices, data, bandwidth, and services. We’ll see the first waves of IoT implementation in diverse pockets of innovation—in our homes, offices, factories, warehouses, and hospitals and in metro infrastructure, transportation, and agriculture.

Today there are Internet-connected smart phones, PCs, tablets, TVs, set-top boxes, gaming consoles, home appliances, security systems, smart meters, personal medical devices, vending machines, and more. Countless consumer items are embedded with RFID tags for inventory management.

In coming waves of IoT development, we’ll see the aggregation of connected devices into truly smart homes, smart factories, smart grids, and smart cities. According to IoT pundits, the Internet will support some 50 billion device nodes by 2020.

Emerging Applications

The IoT vision has existed since the early days of the Internet. The term “Internet of Things” was coined at the turn of this century to describe networks of RFID-equipped objects to ease identification and tracking.

Over the years, the IoT has evolved to include countless devices connected wirelessly to the Internet. These connections can be device-to-person (or vice versa) or M2M. Nowadays, IoT and wireless sensor networks are nearly synonymous.

There is admittedly a “cool factor” to some IoT applications. Consider the profusion of Apple iPhone apps that allow users to remotely monitor and control security, HVAC, and lighting systems with the stroke of a touchscreen.

Verizon, Cisco, and Google also have their own wireless home networking initiatives, and IBM is developing a wide range of IoT solutions. Google garnered lots of media buzz last year over its Android@Home technology demo, using an Android smart phone to control a wireless LED lighting network based on Synapse’s SNAP mesh networking stack.

In general, wireless sensor network applications can be classified around range and data rate:

• Home-area networks tend be relatively short range, supporting medium to high data rates. Application examples include security systems, lighting and HVAC control, and in-home utility monitoring systems.

• Industrial control and factory automation occupy the mid-range with wireless ranges of several kilometers and low-to-medium data rates.

• Metro-area networks and infrastructure applications such as monitoring bridges, railways, and traffic lights require long-range wireless networks, typically over tens of kilometers, with low data rates. Wide-area sensor networks also can be deployed for such unusual applications as tracking cattle and sensing forest fires.

Deployment Issues

Although Wi-Fi is ubiquitous in the home, it’s a relatively expensive and power-hungry protocol with high data rates. Sub-GHz and ZigBee provide a better fit for home automation. There’s also growing industry interest in using IP-based protocol stacks such as 6LoWPAN (IPv6 over Low Power Wireless Personal Area Networks) for home-area networks.

In the mid-range industrial market, we’ll see continued deployment of WirelessHART technology in the 2.4-GHz band for sensor networks. WirelessHART is attractive for industrial control because it supports “five-nines” reliability through a self-healing meshed network architecture.

Long-range wireless sensor networks pose an interesting challenge that will be met with a disruptive technology called TV white space—the analog TV spectrum no longer in use and available in some geographies as free spectrum for wireless data services.

TV white space offers excellent range and data rates scaling to 16 Mbits/s at distances up to 10 km, leading some to call it “Wi-Fi on steroids” for rural broadband. Since it is a free spectrum, TV white space is attractive to both major service providers and IoT technology startups.

A pioneer in TV white space technology to watch is the U.K.-based startup Neul (the French word for “cloud”). Neul’s visionary solution uses white-space channels adjacent to digital broadcasts (unlicensed radio spectrum in the 400- to 800-MHz range) for M2M communications. Target applications include smart metering, asset tracking, point-of-sale, and home automation. TV white-space networking is feasible and happening now, with trials underway in Europe and the U.S.

No one wireless technology will provide a panacea for IoT. Each technology will serve varying application needs based on range, data rate, and cost. It’s also encouraging to know that the basic silicon building blocks for IoT are available today: nano-amp microcontrollers and wireless MCUs, low-power RF ICs, and micro-electromechanical systems (MEMS) and CMOS-based sensing devices to detect pressure, motion, temperature, humidity, and gases.

If a mains power supply is not available for the wireless sensor and if battery use is impractical (often the case in widely dispersed infrastructure applications), energy harvesting technologies can be deployed to provide a sustainable power source.

The proliferation of wireless sensor networks and IoT solutions will drive enormous amounts of data across the Internet, requiring fatter pipes and the expansion of cloud computing. Wireless sensor networks are becoming an integral part of our lives.. This confluence of rapidly emerging wireless, sensing, and signal processing technologies ultimately will enable the Internet of Things.

Mark Downing, vice president of strategy and business development, joined Silicon Laboratories in 2007. Previously, he was chief executive officer and director of Enpirion, a private company offering high efficiency switching dc-dc converters. Before Enpirion, he served as vice president of marketing for Micrel and was responsible for defining strategic direction. He also served as vice president of Marketing at Pericom Semiconductor, where he was responsible for each of Pericom’s four product lines and launching the company in preparation for an IPO. From 1988 to 1997, he served in various strategic planning, marketing, product line management and applications management roles at National Semiconductor. He holds a BS degree in physics from Aston University and an MBA from Open University.