Reputation Based Privacy-Preserving in Location-Dependent Crowdsensing for Vehicles
Huizhu Yang, Chao Jun Yang, Qihang Wu, Weidong Yang · 2024
With the development of the Internet of Vehicles (IoV), the location-dependent crowdsensing has attracted widespread attention. To achieve the optimized corwdsensing task utility, tons of related data should be collected and uploaded to the service provider for task allocation. However, it could introduce potential privacy risks if the untrusted service provider discloses the sensitive data. Thus, the key challenge is how to optimize the overall task utility without directly uploading the sensitive data to the service provider, such as vehicle location, mobile node reputation and task content. To this end, we propose a reputation based privacy-preserving task allocation mechanism for the location-dependent vehicle crowdsensing. Firstly, we design a symmetric key generator to enable privacy-preserving task release between task requester and mobile nodes. Furthermore, we propose a deep reinforcement learning (DRL) based task allocation algorithm with obfuscated location data to optimize the overall task utility. The simulation result reveals that our mechanism could achieve an effective overall task utility while preserving sensitive data from an untrusted service provider.