Learning ensemble strategy for static and dynamic localization in wireless sensor networks

Indoor localization in wireless sensor networks is a challenging task. Static localization and moving target monitoring are addressed using ensemble learning method and received signal strength indicator. The suggested strategy combines several regression trees to have better performance. This solution has been experimentally evaluated using real measurements in an office room. The performance results have been analyzed through a comparison with learning-based localization algorithms currently available in the literature. The analysis shows that the adopted solution is simple in term of computation, accurate and robust to environmental variation.