Research On Data Privacy Preserving Techniques Of Internet Of Things Application

At present, most Internet of things(IOT) applications adopts the logical architecture of IOT cloud platform, IOT intelligent hardware and mobile phone APP. Cloud platform provides the storage and processing of data center for IOT applications, but also results in potential data privacy threats because of untrusted cloud platform. First, the security of the cloud platform itself may affect the reliability of the computation of the data, also cloud administrators may snoop, leak, tamper user data, or even resale data for commercial purposes.Secondly, mass user data of cloud platform forms a clear and large attack surface, and the attacker may be more likely to get the user's privacy data and implement a specific attack. Thirdly, the user is not fully informed of the types,attributes, timing of information collected by cloud platform. Therefore, that how to prevent cloud administrators and attackers from collecting and accessing data without user's authorization is important. Also, it is important that how to ensure the privacy of user data while supporting verification of IOT applications operation. However, at present, there is no literature about the research on the data privacy of Internet of things.This dissertation has studied some key technologies of data privacy preserving and proposed schemes aiming at above data privacy threat of IOT applications. During the design process of the proposed schemes, this dissertation references the latest academic achievements in other fields, such as verifiable computing, searchable encryption, attribute based encryption,oblivious random access machine, et al. The dissertation is committed to constructing schemes for different practical application scene of IOT and testing schemes in practical applications. Specifically, main achievements and innovations of this dissertation are summarized as follows:Firstly, a data privacy preserving scheme of IOT application based verifiable computation is proposed firstly in this dissertation aiming at reliability of cloud platform. The scheme ensures the privacy of user data whiling supporting to verify the correctness of computation. At present, the privacy of user input data in most scheme is not considered. So, the research results and classification in verifiable computation protocol field is illustrated in this dissertation, and the protocol which is suitable to constructing privacy preserving of IOT application is expounded. Specifically speaking, the problem description and the design principle is analyzed, and also the process transforming required computational models from general program is illustrated through concrete examples. Research shows that even the best protocol such as Trinocchio supports only specific scene of data privacy protection, because the performance overhead is still large and not suitable for generic IOT application. Nonetheless, the research in this dissertation plays an important role in understanding and designing the verifiable computing protocol which apply to data privacy preserving of IOT application.Secondly, a data privacy preserving scheme of IOT application based searchable encryption, attributed based encryption and oblivious random access machine technology is proposed for new IOT application. Also the scheme of the protocol process and technical requirement is illustrated. Data privacy preserving is taken into account at the beginning of designing architecture. However, this scheme is suitable for IOT application whose search expression request is not high.Thirdly, a broker based data privacy preserving for IOT application is proposed aiming at unauthorized access of cloud administrators and attackers,also search and authorization scheme across broker is illustrated. Then security and performance of the scheme is analyzed. A performance test shows that performance is not more than 10%. The scheme is transparent to cloud platform and user. At the same time, it needs not to make changes to the cloud platform and supports for full text search. Experiments and performance shows that the proposed scheme has practical significance for the data privacy preserving of existing IOT applications and IOT application whose search expression request is high.Lastly, a fine-grained authorization based data privacy preserving scheme is proposed to solve data authorization of IOT application. Users can not only make full control of the own data, but also can realize flexible data control with this scheme. Data is filtered again before returns to the user. So, user data is filtered by a dual protection, including searchable encryption and access control. The data access can be flexibly realized while ensuring the privacy of data. Experiments and performance show that it got a very high performance among no more than 10 attributes.The security analysis, performance analysis and experimental results show that the proposed scheme of data privacy protection for the IOT application is reasonable in practical scenarios, and important to solve to unauthorized access of cloud administrators and attackers and reliability of cloud platform.