As ubiquitous computing becomes more popular, the need for context-aware applications increases. The context of an application refers to the information that is part of its operating environment. Typically this includes information such as location, activity of people, and the state of other devices. Algorithms and techniques that allow an application to be aware of the location of a device on a map of the environment are a prerequisite for many of these applications.

Many systems over the years have tackled the problem of determining and tracking user position. Examples include GPS, wide-area cellular-based systems, infraredbased systems, magnetic tracking systems, various computer vision systems, physical contact systems, and radio frequency (RF) based systems. Of these, the class of RF-based systems that use an underlying wireless data network, such as the IEEE 802.11 wireless network, to estimate user location has gained attention recently, especially for indoor applications. RF-based techniques provide more ubiquitous coverage than other indoor location determination systems and do not require additional hardware for user location determination, thereby enhancing the value of the wireless data network. However, using a wireless network for location determination has the challenge of dealing with the noisy characteristics of the wireless channel.

Current location determination techniques for the 802.11 wireless networks suffer from these noisy characteristics, leading to coarse grained accuracy. A key feature of current techniques is the dependence on building a radio map for the area of interest and using this radio map to infer the user location. Using the radio map to infer the user location is a computationally intensive process and may consume the scarce energy resource of the mobile units. The Horus system is concerned with developing accurate methods for determining the user location with low computation requirements. Our goal is to build a location determination system that is capable of determining the user position with high accuracy, is light enough to be implemented on energy-constrained devices such as handheld computers, and is scalable to track a large number of users and to be used with large areas.

We identify different causes of the wireless channel variations and we develop techniques to handle these variations. The results show that the Horus system is able to achieve accuracy significantly better than the current WLAN location determination systems. Moreover, the number of operations required to run the algorithm is better than the current systems with more than an order of magnitude.