Privacy Protection Mechanism Of Crowd-sensing Network Based On Blockchain And Trusted Execution Environment

With the popular and development of sensors-containing intelligent terminals,crowdsensing system based on the Internet of Things has become a new paradigm of application.In crowdsensing system,users with smart devices can collect large-scale data in an economical and efficient manner for corwd intelligent extraction and human-centered service delivery.However,most of the existing crowdsensing system applications are based on a centralized structure,and their trust issues are difficult to guarantee user privacy,and users' activity in crowdsensing system will be affected.Existing research mainly combines traditional privacy protection methods such as K-anonymity,obfuscation methods,differential privacy,etc.,but these methods fail to fully consider the integrity of data and the reliability of reward distribution,which leads to the decrease of system availability.At the same time,most of solutions are based on the concept of a trusted third party to complete sensitive operations,and a trusted third party hardly exists in the real environment.Some scholars also consider blockchain technology as a solution to solve the problem of centralized trust.However,the security of sensitive operations still limits the research of decentralized crowdsensing systems.This thesis proposes a privacy protection scheme based on blockchain and trusted execution environment of crowdsensing system,which achieves privacy-preserving and accurately distributing rewards without a trusted third party.This thesis first deploys a trusted execution environment on the task publisher's server to be responsible for sensitive operations such as encryption,decryption and quality assessment,which can ensure the security of the system while getting rid of trusted third parties;Encryption algorithms are used in the interaction between the blockchain and the trusted execution environment to ensure the confidentiality of the data and the disconnection between the user's identity and the sensitive information in the data,while ensuring that only the user who owns the reward can query it from the blockchain and decrypt;Finally,it proposes three methods to solve the possible association attacks and user security attacks in the system with K-anonymity,uniform data size,and private key signature verification.In this thesis,two comparative experiments of traditional centralized server and Hyperledger Fabric,ordinary environment and trusted execution environment are used to measure the impact of the blockchain and trusted execution environment on the overall system performance.