LandLoc: Landmark-based User Location

The growing interest in location-aware computing has spurred the development of systems and technologies to locate users, both in indoor environments and the outdoors. These include the global positioning system (GPS), the forthcoming mobile phone based wireless E-911 system, and several indoor location systems. These systems depend on distributed infrastructure support to determine user location, which makes them susceptible to the non-availability or temporary failure of the infrastructure. The cost of the dedicated infrastructure and the potential threat to user privacy in some cases are also drawbacks. In this paper, we present a novel approach to the user-location problem that requires little or no infrastructural support. We call this approach landmark-based location or LandLoc for short. The basic idea is that the user identifies landmarks in his or her vicinity and feeds this information into a mobile computer. The system compares the landmark information against a 3D topographical model of the physical world, thereby narrowing in on the user’s location. We discuss the technical challenges presented by LandLoc. We do not claim that the LandLoc approach is superior to existing schemes in all respects. Rather we present it as a promising research direction that offers several advantages and may yield a solution complementary to existing systems.