Fingerprint Localization with Adaptive Area Search

A new set of complete fingerprints positioning algorithm is proposed in this letter. Firstly, a new criterion that converts the RSS difference between reference points and test point into a relative ranking is proposed to select one reference point with highest similarity to the test point. Secondly, {k} reference points with higher similarity are adaptively selected by this criterion and an adaptive area search algorithm based on region growing, then the inverse of each reference point ranking is adopted as their own weight. Finally, the coordinates of {k} reference points are weighted average to obtain the positioning result. Extensive experimental results show that the proposed algorithm can effectively improve the positioning accuracy.