Research On Key Technologies Of Homomorphic Privacy Preservation In IoT

The Internet of things(IoT)is the network connecting people and people,people and objects,objects and objects with information sensing equipments according to protocols.There are a lot of security threats in the applications provided by IoT.A great challenge to IoT is how to protect privacy or sensitive information from disclosing during application process of IoT.It is also the key to the large-scale applications of IoT.Researchers at home and aboard have done some work on privacy preservation in IoT and there is no systemic research on it yet.Based on analyzing of the definition,characteristics and the network architecture of IoT,this dissertation summarizes privacy problems and current privacy-preserving methods by discussing privacy threats faced by IoT.The object of this dissertation is to study on key technologies of privacy preservation in IoT and proposes privacy-preserving methods based on homomorphic encryption for data privacy protection,location privacy protection and secure authentication in IoT.Major contributions of this dissertation can be summarized as follows:1)It proposes privacy-preserving methods based on the homomorphic Elliptic Curves Cryptograph(ECC)algorithm,which can be used for data aggregation privacy protection in wireless sensor networks,distributed-computing and distributed-storing.The homomorphic characteristic of the method ensures that users can process sensitive data without revealing data information to other users,illegal user and even the aggregation center.ECC homomorphic method can also be used for cipher-processing in distributed-store.The ECC algorithm is suitable for resource constrained application in IoT,due to its characteristics of low computation consumption and high security.2)It proposes homomorphic privacy-preserving methods based on homomorphic hash function for secure authentications of IoT and privacy-preserving during integrity verification.The secure mutual authentication methods based on homomorphic hash function and ECC can be used for mutual authentications in WSN,distributed systems and RFID systems.The hash function has characteristics of low resource-consumption and high security,which is suitable for storing and computing constrained applications.The hash method can achieve mutual authentications based on one hash computing by setting computing location 1n and2 n and solves ID updating problems by setting new and old IDs.In the hash and ECC method,the secret key and the hash function is used separately for authenticating different users.The privacy-preserving method based on homomorphic hash and ECC signature can protect data privacy during integrity verifications fulfilled by the third party.3)It proposes a privacy-preserving method based on secure routing protocols,which can be used for location privacy protection in WSN.The method includes three parts.At first,the source sensor node slices data into pieces and encrypts each slice with ECC to ensure data security during data transferring.Then the source node sends slices to different phantom nodes based on the direction classification,which can achieve better protection for the location of source node compared to current phantom routing protocols.At last,the phantom nodes send slices to the sink node separately based on the clustering routing protocol and P2 P protocol,which has lower communication consumption and longer lifetime compared with flat routing.The sink node decrypts and recombines slices to original data.The homomorphic property of the method ensures the security during data aggregation in cluster heads.4)It proposes a method based on network coding,which can be used for both data and location privacy protection in WSN.The method can protect location privacy of source nodes from flow analysis attacks and flow-tracing threats,due to the mixing coding mechanism of network coding.Besides,the method encrypts encoding vectors with ECC homomorphic encryption,which ensures data security from intermediate re-encoding nodes.