Passive localization methods based on LTE sniffer

Abstract

The LTE Sniffer is a critical tool for analyzing LTE networks, and the addition of uplink channel support in open-source sniffers has spurred research into passive localization based on LTE networks. Previous studies on network-based localization faced several limitations, including overly simplified experimental scenarios, the absence of commercial base stations, limited localization accuracy, and a lack of targeted improvements in localization algorithms. To address these issues, we extended the localization scenario to include commercial base stations and made significant advancements at the algorithmic level. By combining the Taylor algorithm with Z-score-based filtering methods, we improved the passive localization accuracy using solely commercial LTE signals to approximately 10 meters, achieving a practical and deployable level of performance. This dissertation begins with an overview of prior research on which our experiments are based, followed by a detailed analysis of the physical layer information transmission necessary for the experiment. Subsequently, we modeled the experimental scenarios using the TDoA-based approach and derived the formulas for different localization algorithms, such as the spherical intersection method and the Taylor algorithm. Finally, we implemented these methods through experiments and code execution, validating their effectiveness.