Fusing Measured and Simulated Data for Radio Map Construction in Localization Applications

Abstract

Most localization algorithms can be classified into the geometrical approach, where the location of the target is determined geometrically using the estimated range of the target or the angle of arrival (AOA) of the target’s signal at receiver sensors. Range and AOA estimation require the line-of-sight (LOS) between the target and sensors, which is unlikely in a dense urban scenario where the antenna heights are much lower than the building heights. It is expected that fingerprinting approaches, which do not require LOS, achieve higher localization accuracy than geometrical ones.