Research On Anonymous Communication Model And Key Technologies In The Internet Of Things

As the wireless communication technology (such as short-distance wireless communication, Wi-Fi, etc.) are widely used, Internet will be extended to the real world. Internet of Things (IoT) becomes the important infrastructure of the next generation network. However, the booming of Internet of Things changes the way of humans’interaction, but also inevitably bring a huge security and privacy risks, such as identity privacy, communication privacy, service privacy and so on. Aiming these privacy issues, this paper makes a research on IoT communication system, builds an anonymous communication model toward Internet of Things environments, and carries out in-depth research on anti-attack technology of IoT anonymous communication model and privacy protection technology of sharing and social discovering service.First, this thesis reviews existing IoT communication models and architecture reference model (ARM). On this basis, the IoT communication features are analyzed and given a formal definition. Also, a review of the existing privacy-related technologies of IoT environment is done, including privacy protection model, identity-based privacy protection technology, location-based privacy protection technology and routing-based privacy protection technology. And their application environment and technology features are also analyzed.Next, for solving the communication privacy, the property of IoT is considered, such as a large number of nodes, limited computation and storage resources, as well as frequently joining and exiting. A node anonymous communication model which orients the IoT environment is proposed. This model consists of the sender, recipient, self-organization anonymous forwarding network and security transmission mechanism. Its core parts are group-based anonymous forwarding network, security transmission mechanism based on network coding and homorphic encryption function. In this model, pseudomyms is used to implement sender anonymity; broadcasting is used to achieve recipient anonymity; the group-based anonymous forwarding network between sender and recipient implements communication relationship anonymity; security transmission mechanism based on network coding and homorphic encryption function ensures the confidentiality of data transmission. The theoretical proof and experimental results show that the model has good anonymity and anti-attack ability. Meanwhile, it reduces the number of key agreement and encryption, and reduce the computation cost of nodes. Moreover, it has good tolerance and scalability without a central server.Then, in order to improve the anti-attack capability of proposed node anonymous communication model, for Sybil attack that is a type of important routing attack, A Sybil attack detection and prevention mechanism based on community detection are proposed. The Sybil attack detection algorithm consists of trust network construction, community detection, attack edge detection and Sybil node location. The trust network is created by metricing the trust relationship between nodes. Sybil communities and honest communities are detected in the trust network according to the node trust relationship. Then, attack edges are detected by edges between Sybil community and honest community. Above all, Sybil nodes are located using discovered attack edges. The experimental results show the Sybil node detection method based on community detection has a better detection rate, whenever the number of honest nodes is dominant or the number of malicious nodes is dominant. The correct rate of Sybil node detection algorithm is relatively stable, and the Sybil based on trust routing mechanism can effectively reduce the impact of Sybil attacks on packet delivery ratio.Finally, for the identity privacy leaking issue in the sharing and social discovering service of Internet of things, an attribute based anonymous matching protocol and its enhanced protocol are proposed. The attribute based anonymous matching protocol uses a node attribute set as identity. The identity anonymity is achieved by changing each attribute to an attribute key. For the inference attack on sensitive information in the proposed anonymous matching protocol, its enhanced privacy matching protocol is proposed, which further enhances the node identity privacy against inference attack. The theory proved that the above protocols can effectively protect the identity privacy in sharing and social discovering service in Internet of things.In all, the privacy protection is an important part of IoT research. In this thesis, IoT node privacy issues are being focused on, such as lack of an anonymous communication model, poor resistance while attacking, and sensitive information concealing in discover service. Based on the existing work, we analyzed their disadvantages, and built the IoT node anonymous communication model. Further, the anti-attack technology and privacy preserving technology in the model are studied in depth. Our research contributions provide an important theoretical model and technical schemes for IoT privacy problems, which would effectively improve the level of privacy and enhance anti-attack ability.