Mobile Technology

KAIST develops new Wi-Fi indoor positioning system

KAIST develops new Wi-Fi indoor positioning system
KAIST researchers have developed an indoor positioning system based on "Wi-Fi fingerprints" from mobile devices and signal strengths from APs
KAIST researchers have developed an indoor positioning system based on "Wi-Fi fingerprints" from mobile devices and signal strengths from APs
View 2 Images
KAIST researchers have developed an indoor positioning system based on "Wi-Fi fingerprints" from mobile devices and signal strengths from APs
1/2
KAIST researchers have developed an indoor positioning system based on "Wi-Fi fingerprints" from mobile devices and signal strengths from APs
To test the method, the research team collected 7,000 Wi-Fi fingerprints at 400 access points in four selected areas in Korea
2/2
To test the method, the research team collected 7,000 Wi-Fi fingerprints at 400 access points in four selected areas in Korea

Researchers at the Korean Advanced Institute of Science and Technology (KAIST) have developed a new indoor positioning system that makes it possible to build a Wi-Fi radio map that does not require GPS signals. It's claimed to be an improvement on Wi-Fi Positioning Systems, that rely on both GPS and Wi-Fi signals.

The system consists of a set of Wi-Fi LAN signals captured by a mobile device, and the measurements of received Wi-Fi signal strengths (RSSs) from access points surrounding the device. The Wi-Fi radio map shows the RSSs of Wi-Fi access points at different locations in a given environment, so each "Wi-Fi fingerprint" on the radio map is linked to data on location.

For their research, the KAIST team collected fingerprints from users' smartphones every 30 minutes for a week, through modules embedded in mobile platforms, utilities or applications. After that, they analyzed the characteristics of the collected fingerprints based on the users’ home and office addresses as location references. Then they classified the fingerprints collected from the phones according to one of those two locations.

Using Google’s geocoding, Prof. Dong-Soo Han converted each home and office address into geographic coordinates to obtain the location of the collected fingerprints. The Wi-Fi radio map has both the fingerprints and coordinates, whereby the location of the phones can be identified or tracked.

In order to test the method, the research team collected 7,000 Wi-Fi fingerprints at 400 access points in each of four selected areas in Korea. They were careful to choose a mix of commercial and residential locations to collect data for the Wi-Fi radio map. The researchers realized that the more data they collected, the more accurately they could determine location. With a collection rate of 50 percent (that is, with every other house or office in the area covered), the margin of error was not more than 10 meters (32.8 ft).

The KAIST researchers want a non-GPS alternative because GPS signals do not work well in indoor spaces or between high-rise buildings, as they lose two-thirds of their accuracy in vertical configurations – they need a clear view of the sky to communicate with satellites.

The researchers still need to deal with privacy protection issues before launching the method commercially, but they believe that once the address-based radio map is fully developed for commercial use, it will be useful for a variety of applications. These could include emergency rescue or tracking of lost cell phones, missing persons, or survivors during fires.

Source: KAIST

1 comment
1 comment
Jon
Sensors and correspondence advancements are used by indoor positioning systems (IPS) to locate items in interior surroundings. There have been numerous previous studies on indoor positioning frameworks; however, many of them lack a strong characterization scheme that would fundamentally plan a broad field, such as IPS, or exclude a few key innovations or have a limited viewpoint; finally, studies quickly become outdated in a field as unique as IPS. You should choose for basic quality when it comes to out-positioning. UbiTrack is a good place to start.