Single-site localization via maximum discrimination multipath fingerprinting

A novel approach to single-site localization based on maximum discrimination multipath fingerprinting is presented. In contrast to the existing approach, which extracts each fingerprint only from the data of that location, the new approach uses also the data of all the other locations in the database, and leverages it to extract a fingerprint that is as different as possible from the other fingerprints in the database. The performance of this approach, validated with both simulated and real data, is superior to the existing approach, demonstrating single-site localization accuracy of 1 m in typical indoor environments. The new approach has also a lower computational complexity.