Ultimate precision bound and comparison study of dynamic indoor localization systems

Many location determination systems have been developed lately due to growing demands on location based applications and services, however, the performance of most systems is not well understood because a typical indoor system would depend on a lot of factors and the conditions under which reported tests were carried out could be very different. In this work, we analyze theoretical localization limits and propose a precision bound as a benchmark standard to evaluate the performance of indoor systems under different configuration settings and environmental dynamics. We examine a set of localization mechanisms using data sets measured from an underground floor plan. We compare the performance of these systems and analyze their unique responses to the change of system configurations (such as sniffer deployment methods and reference measurements) and localization processing techniques, we found that database-based indoor systems generally perform better than the distance-based systems. From the comparison study, we observed that while it appears difficult to interpret the relationship between the performance of a localization system and various environmental and system factors, the achievable precision of a location determination system deployed at a given building is bounded and this bound is predictable. Thus this research provides a valuable criterion for the evaluation of location determination systems.