Deterministic and Probabilistic Fingerprint-based Localization utilizing RSS and Phase-difference

摘要

Fingerprint-based localization is expected to be superior than conventional techniques in non-line-of-sight (NLOS) environments. In this paper, we perform a detailed comparison between these deterministic and probabilistic approaches, focusing on the popular received signal strength (RSS) fingerprint, and our recently proposed phase-difference fingerprint. We also proposed a hybrid algorithm combining these two fingerprints using kernel densities. Simulation results indicate the superiority of hybrid algorithms. We also highlighted some issues when dealing with multipath fading in street canyons at low frequencies.