Development of security and Privacy Techniques in mobile Crowdsensing

Abstract

Mobile CrowdSensing (MCS) embodies a dynamic paradigm leveraging the pervasive nature newline of mobile devices equipped with advanced sensors to collect environmental data for collective newline analysis and processing in pursuit of shared objectives. This thesis extensively explores the newline intricate security and privacy challenges inherent in MCS, placing a strong emphasis on safe newlineguarding user privacy while ensuring accountability, implementing effective data obfuscation newline techniques, upholding data integrity and establishing mechanisms for user reputation manage newlinement. newline In the architecture of Mobile CrowdSensing (MCS), where mobile devices are not centrally newline owned, participants are inclined to engage in MCS campaigns only if their privacy particu newlinelarly that of their mobile devices is safeguarded. However, the inherent openness of the MCS newline platform also creates vulnerabilities like allowing participants to potentially submit false or de newlineceptive data. Consequently, it becomes imperative for the platform to detect and address such newline instances of malicious behavior, thereby ensuring accountability among participants. This may newline entail implementing measures to identify and take appropriate action against those found to newline be engaging in misconduct such as imposing penalties or removing them from the platform. newline To this end, the thesis introduces a novel architecture named A-Anonysense which seamlessly newline integrates user accountability while preserving privacy. By leveraging distributed threshold newline cryptography, this architecture ensures user anonymity while maintaining the essential aspect newline of accountability. newline The widespread availability of sensitive user data within MCS environments underscores newline the need for nuanced approaches to address privacy concerns effectively. The thesis proposes newline a Privacy-Aware Missing Data Inference (PAMDI) approach to safeguard user data privacy newline during data inference processes.