An experimental evaluation of agent-based indoor localization

This paper discusses the possibility of accurately estimating the position of smart appliances in known indoor environments using agent technology. The proposed localization approach does not require specific infrastructures, nor it requires special on-board sensors, because it uses only WiFi signaling, which can be considered ubiquitous. Localization is performed using range estimates from fixed access points of the WiFi network with known positions. In order to derive the experimental results shown in the paper, a specific range-based localization algorithm, the Two-Stage Maximum-Likelihood localization algorithm, is considered. Results presented in the last part of the paper show that a proper processing on the WiFi-based range estimates allows obtaining position estimates which can be considered sufficiently accurate for many applications