earticle

영어

The deployment of a fingerprint positioning method consists of an offline phase and a real-time phase. During the offline phase, a fingerprint database is built. This process is tedious and time consuming. Even so, the majority of WiFi-based indoor positioning methods implemented in practical indoor location–based service (ILBS) systems are the fingerprint method, because this method is accurate enough, whereas other WiFi-based indoor positioning methods are too inaccurate. During the real-time phase, the fingerprint method obtains a test fingerprint (a set of WiFi signals collected at that moment). Then, for each fingerprint, fi , in the fingerprint database, it compares fi with the test fingerprint in order to find the most similar fi . This paper introduces a novel method of comparing two fingerprints in order to improve the accuracy of the fingerprint method. This method assigns weights to received signal strength indications (RSSIs) based on the variance of the RSSIs.