Fault tolerant based group key Management with secure and memory Efficient data deduplication using inlineBlock matching protocol for Cloud storage

Abstract

Cloud is an internet-based computing technology, which provides software, platform, and storage services to customers. With the substantial increase in usage of Cloud Computing, the need for security to protect people s information and efficient utilization of storage is an urgent problem to solve. This thesis focuses on group key management in the cloud using centralized and contributory key management with the help of clustered structure. The key server plays a major role in the centralized key management scheme to generate keys for both users as well as groups. The proposed work consists of one central server and multiple sub-servers, where each sub-server is designed in such a way as to manage different groups. To achieve fault tolerance in the system, each group is replicated into multiple different sub-servers. The communication between the groups is done through a central server. Internal attacks are avoided by maintaining a separate group key for each group. Here, the costly rekeying process of new user joining has been eliminated by maintaining a file access list for each file. With the inclusion of file access lists, the server does not suffer, even when there is a leakage of the group key after the user leaves. To avoid single-point failure, a contributory key management scheme is also proposed. In this scheme, each group member in the group is responsible for generating the new group key for the group where the user leaves. The major outcomes of the work are the generation of group keys and secured file sharing. The system can run without any inconsistencies even if the central server is not available. The performance of the proposed work is evaluated in terms of comparing the computation and communication time of the group key with existing works.. newline