AI / ML BASED METHOD FOR LOCATION OF USER EQUIPMENT USING FINGERPRINTS

Abstract

This description presents a method and system for precise localization of a target user equipment (UE) (210) using signal fingerprinting and machine learning (ML) techniques. The method is performed by a base station (BS) (100) within a coverage area (300) and involves the transmission of training downlink reference signals (DL-RS) (400) along combinations of takeoff angle and transmission power levels to perform directional scanning; the reception and measurement of corresponding reflected DL-RS (R-DL-RS); and the acquisition of uplink reference signals (UL-RS) (500) and associated location information from numerous UEs (200). These measurements are stored and used to train a machine learning model to match signal characteristics with known UE (200) locations. Once the model is trained, it provides real-time location estimation of the target UE (210) based on the newly acquired R-DL-RS and UL-RS (500) measurements. Figure 1