A Double Obfuscation Approach for Protecting the Privacy of IoT Location based Applications

• Published in: IEEE access 2020-01, Vol.8, p.1-1
• Main Authors: Albouq, Sami S., Sen, Adnan Ahmed Abi, Namoun, Abdallah, Bahbouh, Nour Mahmoud, Alkhodre, Ahmad B., Alshanqiti, Abdullah
• Format: Article
• Language: English
• Subjects: Accuracy; Algorithms; Caching; Cloud computing; Connected Vehicles; Data privacy; Edge computing; Internet of Things; Location based Services; Nodes; Obfuscation; Privacy; Queries; Response time; Security; Smart cars; System performance; Trusted third parties; Trusted Third Party
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2020.3009200

Connected vehicles and smart cars have become highly reliant on location-based services (i.e. LBS) to provide accurate, personalized and intelligent services. However, location-based services have endangered its users to considerable risks concerning the privacy and security of users' personal data. Although existing research provides a myriad of methods to improve and protect user privacy in LBS applications, most of these methods are concerned with handling static queries and non-mobile objects only. Moreover, various issues and challenges still persist with regards to the need to trust third parties, overloading of the user, and low accuracy of the returned results. This paper contributes a Double Obfuscation Approach (referred to as DOA) that applies two phases of obfuscation consecutively whilst integrating two differing privacy protection approaches, namely Obfuscation and Trusted Third Party, and two techniques, namely fog caching technology and mix zone. In essence, the DOA obfuscates and hides the identity and location of its users using the fog nodes, which operate as a trusted third party (TTP), and without the need to reveal the identity of the users or trust the cooperating nodes. Moreover, this paper presents a DOA algorithm that improves the overall user privacy and system performance using the fog nodes, which split the responses of each query into five parts, thus reducing the processing time of the results by the user and enhancing the overall accuracy where the user directly selects the most suitable parts based on his current location. Overall, the hybrid DOA approach empowers the users of connected vehicle applications to protect their privacy through an algorithm that caters for the dynamic nature of user queries and mobility of objects. The results of our comparative simulations against well-known hybrid privacy protection methods demonstrate the superiority of the proposed Double Obfuscation Approach especially with respect to user privacy whilst maintaining a nominal overhead on the user, reduced response time and high accuracy of the obtained results.