A Machine Learning Approach for Hierarchical Localization Based on Multipath MIMO Fingerprints

Localization based on channel state information (CSI) fingerprints in multiple-input multiple-output (MIMO) systems is one of the most promising positioning technologies. In this letter, a machine learning approach is developed for hierarchical localization exploiting multipath MIMO fingerprints. By using multipath MIMO CSI in the time domain, instead in the frequency domain, our hierarchical scheme uses coarse-to-fine process and can achieve accurate positioning, which is demonstrated by simulation results.