Robotics and eSafety Research Group

University of Alcalá

  • Increase font size
  • Default font size
  • Decrease font size

WiFi RTLS Description

Traditionally, global localization has been carried out through GPS, which provides accurate localization when working outdoors. Hence, GPS has become the main technology for positioning in outdoor environments. Encouraged by GPS’s accuracy outdoors and due to the fact that almost every new smartphones and tablets have built-in GPS receivers, most of the currently existing Location Based Services (LBSs) are oriented to outdoor environments. However, in indoor environments, like in hospitals, malls and museums, LBSs are of equal interest in a wide range of personal and commercial applications.

This kind of applications requires accurate indoor localization for the strategic planning guidance to the final target. Indoor localization techniques require the use of other technologies such as Ultrasound, Infrared or Radio Frequency (RF). RF indoor localization based on WiFi is arising as the most popular one. This is probably due to the advantages of using WiFi for indoor localization: WiFi Access Points (APs) are deployed in almost every building and measuring the WiFi signal is free of charge even for private networks. This fact allows developing a localization system based on WiFi without modifying the environment. Moreover, almost every device is already equipped with a WiFi interface, and no special requirements are needed to perform localization. Hence, widely used devices such as smartphones, tablets or laptops can benefit from indoor LBSs using WiFi.

Unfortunately, using WiFi for indoor localization also has some disadvantages due to the fact that WiFi was not originally intended to be used as a localization technology. The Received Signal Strength (RSS), which is used for localization, decays logarithmically on free space. However, in indoors it is strongly dependent on the building structure and some other non-desired effects. Most of these effects are due to the multipath effect, obstacles and the small scale effect which make the RSS a complex function of the distance. In addition, most of the commercial devices equipped with WiFi technology use 802.11b/g standards which work at 2.4 GHz, very close to the water resonance frequency.

WiFi technology at 2.4GHz works in a free band frequency, where some other devices such as Bluetooth work, and it makes the WiFi RSS a noisy signal. Another important issue is the absorption of part of the signal by people (mainly composed by water) moving around in the environment, which significantly diminishes RSS. In addition, the small scale variations cause a chaotic RSS variation when the WiFi device moves distances under the range of the wavelength (12.5cm for 2.4GHz). This effect makes very difficult to estimate the correct location because small variations in the position can lead to high variations in the RSS.

Moreover, since WiFi networks are deployed with the goal of maximizing connectivity and disregarding localization tasks, there are usually a high number of APs distributed over the environment increasing the so-called co-channel interferences, which cause high variations in the RSS from the APs.

Given the diversity of WiFi localization methods published in the literature, different classifications of them are possible. Even though classification of some of the methods is not evident and there is a certain degree of overlapping between works, they can be classified in terms of the algorithms that are used to solve the localization problem (trilateration based on propagation models, fingerprint methods with machine learning, probabilistic approaches, etc).

While the propagation model based algorithms are easy to deploy, the accuracy usually are the worse one. Respect to the second group, fingerprinting methods, we can find a very good accuracy but also a hard training or calibration stage to build the radio map. In addition, environments can change over time. Sometimes it may be relatively slow, such as the structural changes in buildings. But it can be faster, such as the change of a door status or the location of furniture items. Both of these changes can affect the WiFi RSS and then affect the localization system. Therefore, there is a need to use up to date maps of the environment and signal in order to keep an accurate localization.

The third one, the probabilistic approach, provides the most accurate solution but using a well modelled observation and action functions.