Fingerprinting-Based Localization in 5G: a Privacy and Security Perspective

Fingerprinting-based localization is increasingly adopted in 5G to achieve sub-meter positioning accuracy by exploiting radio features such as channel state information (CSI), received signal strength (RSS), and radio frequency (RF) emissions. However, these advances expose users to significant privacy and security risks, including re-identification, session linkage, and continuous tracking, which remain underexplored in existing surveys. This paper provides a comprehensive analysis of fingerprinting-based localization in 5G with a focus on privacy and security. We categorize fingerprinting methods including signal-based, hybrid, machine learning, deep learning, and quantum-enhanced approaches, and compare their accuracy, robustness, computational cost, and inherent privacy resilience. We also introduce a taxonomy of privacy attack vectors such as session linkage, impersonation, correlation, and inference-based tracking, and map them to countermeasures such as pseudonym schemes, federated learning, differential privacy, and blockchain-based authentication. Finally, we highlight research gaps and future directions, including quantum fingerprinting feasibility, integration of privacy-by-design, and the extension of fingerprinting protections to 6G authentication and key agreement (AKA) protocols. Our study reframes fingerprinting research around privacy-first principles to guide the design of secure and trustworthy next-generation localization systems.