Enabling Reputation and Trust in Privacy-Preserving Mobile Sensing

Mobile sensing is becoming a popular paradigm to collect information from and outsource tasks to mobile users. These applications deal with lot of personal information, e. g., identity and location. Therefore, we need to pay a deeper attention to privacy and anonymity. However, the knowledge of the data source is desired to evaluate the trustworthiness of the sensing data. Anonymity and trust become two conflicting objectives in mobile sensing. In this paper, we propose ARTSense, a framework to solve the problem of "trust without identity" in mobile sensing. Our solution consists of a privacy-preserving provenance model, a data trust assessment scheme and an anonymous reputation management protocol. In contrast to other recent solutions, our scheme does not require a trusted third party and both positive and negative reputation updates can be enforced. In the trust assessment, we consider contextual factors that dynamically affects the trustworthiness of the sensing data as well as the mutual support and conflict among data from difference sources. Security analysis shows that ARTSense achieves our desired anonymity and security goals. Our prototype implementation on Android demonstrates that ARTSense incurs minimal computation overhead on mobile devices, and simulation results justify that ARTSense captures the trust of information and reputation of participants accurately.