Intelligent type 2 fuzzy-based mobile application for indoor geolocalization

Geolocalization is a keyword for the emerging location-based applications. This paper presents a new architecture of a geolocalization Android application based on artificial intelligence concept. Our approach considers two contributions: In the learning stage, the system provides an interval-type 2 Fuzzy logic (IT2 FL) processing of the collected radio signal strength (RSS) fingerprints from the Wi-Fi access points. The second contribution is on the output side, a fuzzy location indicator (FLI) is defined to characterize the map zones and rooms. FLIs are type 1 fuzzy sets that will ensure linguistic localization. Then, a Wang-Mendel algorithm is applied to generate the rule base mapping the RSS and their corresponding FLI. For the online stage, the algorithm is implemented and tested using an indoor localization mobile application through the Cynapsys company premises. Using this intelligent localization algorithm based on (IT2 FL), the application has proved better positioning accuracy then classical Wi-Fi fingerprinting systems in zone level and presents ergonomic positioning process through the linguistic learning.