Securing 5G Positioning and its Services with Privacy Preservation

Different from the global positioning system (GPS), the positioning in the fifth-generation (5G) cellular networks is measured through nearby access nodes and processed at the cloud/edge/fog devices. Owing to the availability of high-quality measurements and outsourced computation, the 5G positioning promises high precision, high reliability, wide coverage and low power consumption. The 5G positioning ecosystem relates to 5G positioning and its services. There are four main stakeholders in the ecosystem: location information service provider (LISP), location-based service provider (LBSP), user equipment (UE) with a 5G connection and the location information collaborator (LIC).Focusing on 5G positioning and its services, the present dissertation aims to investigate and resolve the problems in the area of security, privacy and integrity. (1) The security of 5G positioning is threatened by various attacks from signal jamming and counterfeiting to malicious or untrusted devices and users. For solving the security problem in 5G positioning, a framework composed of three modules is proposed to defend against jamming and collusion attacks. (2) To prevent the privacy violation in outsourced 5G positioning computation, two protocols (Pub-pos and Pri-pos) with flexible privacy selection are proposed. (3) Also in the case of outsourced 5G positioning services, we apply an integrity check method by creating a backdoor in an outsourced positioning model based on machine learning. (4) LIC facilitates the position verification by interacting with nearby UEs through distributed device-to-device (D2D) communication. However, the position of private LIC is leaked in the position verification process. To solve this problem, we propose a privacy preservation scheme implemented with the double order-preserving encryption (OPE) and a coordinate-based verification method. Experimental results show significant performance improvement. (5) LBS provision is conducted between the UE and LBSP. For privacy protection, the UE wants to hide its position information and LBSP wants to protect its database from any unauthorized access. However, it is challenging to support a variety of LBS queries and meet the low latency requirement in LBS provision based on 5G positioning, especially when both UE position privacy and LBSP data privacy should be protected at the same time. We propose two protocols, based on exact and fuzzy kNN queries, to achieve mutual privacy preservation, flexible keyword search and low latency. All the proposed schemes are evaluated with simulations or real-world datasets. The results demonstrate the improvement or the trade-off among security, privacy, integrity and overhead. It is expected that this dissertation can further advance secure and privacy-preserving 5G positioning and its services