While a global positioning system (GPS) is the solution of choice for an outdoor location, it’s not effective for locating people or objects indoors, because walls and ceilings block the satellite signals it uses. To this end, Bahl and his colleagues set out to develop a successful indoor counterpart to GPS, and RADAR—the world’s very first Wi-Fi–based indoor location positioning system (LPS)—was born.1

"When I first looked at the issue of an indoor LPS," says Bahl, "I realized that building a separate infrastructure was neither practical nor cost-effective for businesses. So I asked the question: Can we use the wireless LAN infrastructure that already exists inside many businesses?"

With RADAR, Bahl answered his own question with a definitive Yes. Through the use of advanced algorithms, RADAR brings new value to infrastructure-based wireless radio frequency (RF) local area networks (LANs), allowing business and organizations to use existing hardware for much more than data connectivity. These algorithms enable devices to locate themselves inside a Wi-Fi network. Alternatively, operators of indoor Wi-Fi networks can track and locate mobile users or devices in real time for purposes as simple as finding a nearby computing resource such as a printer in an office building, guiding a user through an unfamiliar building, or locating high-value items such as specialized equipment in a hospital.

As a user with a mobile device (for example, a mobile phone or a portable computer) wanders the hallways of a building that supports a wireless RF LAN, the signal strengths that the device measures from the building’s wireless access points (APs) vary depending on where it is. RADAR compares the signal strength from these APs against a database of previously-collected signal strengths and building locations where the strengths were measured, to locate the user’s position on the map. RADAR is completely software-based, so no new hardware beyond the existing APs needs to be installed.

Commenting on the intellectual property (IP) held by Microsoft Research in the area of RTLS, Bahl says, "The key insight that made RADAR work was that we had to exploit signal strength at the receiver and transmitter to determine locations of individuals. This concept is fundamental to indoor location/Wi-Fi based systems."

You Are Here—Zeroing In on Locations

To help RADAR reach new performance level highs in indoor scenarios, Microsoft Researcher John Krumm set about making some improvements to the system’s basic algorithms, which resulted in the ability to locate a wireless device to within one and a half meters’ accuracy, on average.

Krumm’s early work focused on minimizing the amount of RF profiling required to implement indoor LPSs, thereby reducing the calibration effort. RF profiling is the activity of mapping out environmental objects and different signal strengths throughout an indoor area. Krumm also developed a system called LOCADIO, which explicitly acknowledges that there is some noise present when you measure Wi-Fi signals. The noise causes the signals measured from a static location to vary over time. To account for these variances, LOCADIO uses probabilistic modeling to improve the final location answer that is returned.

LOCADIO also takes into account motion models and models of feasible paths to improve the accuracy of RADAR. For example, it acknowledges and exploits walking constraints within the floor plan of a building (people can’t walk through walls), and also makes RADAR cognizant of walking speeds (people can only move so fast), so that even if it looks like the user followed a feasible path, if the speed the user traveled looks unreasonable as compared to normal walking speeds, RADAR will make appropriate adjustments to improve the final calculations. To this end, LOCADIO can also determine whether or not a user is in motion.

Discussing the importance and accessibility of RADAR, Krumm says, "Although we used Wi-Fi radio frequency in these projects, the basic concept of these projects applies to any kind of radio technology—from cell towers, to AM/FM radio, to television, to Bluetooth radios. Any kind of signal strength that is measurable as broadcast from static base stations—indoors and outdoors."

Taking Advantage of Microsoft RTLS

The field of RTLS has entered a fast growth phase since RADAR was introduced, and, according to the Yankee Group2, is expected to reach U.S.$1.6 billion in revenues by the year 2010. While there are several alternative approaches, RADAR appears to be the most economically viable indoor RTLS solution because it takes advantage of existing technology investments to provide new capabilities that companies or organizations can use to reap additional value from existing systems.

Microsoft currently holds more than 40 U.S. patents (both issued and pending), some filed as early as 1999, and several corresponding patents in other countries that encompass the various components of an end-to-end RTLS solution. Microsoft launched an IP Licensing program in December 2003 to provide access to the opportunities created by the company’s significant investments in research and development (R&D). As a result, this IP portfolio, which is composed of patents, software code, and know-how, has recently become available for licensing to third parties under commercially reasonable terms.

"Intellectual property generated by Microsoft can turn into opportunity for others," says Louis Carbonneau, General Manager for the IP Licensing Group at Microsoft. "Our research labs generate more IP than our product groups can absorb. Our goal is to see these technologies deployed; we license them to other companies, in order to help those organizations enter new markets, bring products to market faster, and expand current lines of business."

Microsoft is looking for strategic partners who are interested in expanding the capabilities of wireless technology and bringing powerful new solutions to ubiquity. Companies interested in the Microsoft RTLS IP portfolio can visit the Microsoft Intellectual Property Licensing Web site or send e-mail to Contact IPL.