Research On Assured Deletion Algorithm For Cloud Shared Data

Cloud storage provides the convenience of data management and data sharing for data owners to store their data on remote cloud servers.Since the data owner loses direct control over the data,the assured deletion of data becomes a critical and challenging issue in cloud storage as how to delete the data stored on the remote cloud storage space and ensure that the data deletion operation is actually performed.However,the characteristics of cloud services,such as multiuser and virtualization,make the data that is considered to be deleted but still has the risk of leakage.To this end,this paper proposes two data deletion algorithms to securely and efficiently delete data stored on cloud servers based on the study of the model of data deletion and the security threats faced by the model.The main work is as follows:(1)Considering the existing attribute-based data deletion algorithms,they ignore the excessive overhead of deletion operations due to differences in access control structures and the problem that data can still be accessed after deletion due to attribute revocation omission.To this end,this paper proposes a data assured deletion algorithm based on attribute combination to reduce the cost of data deletion and protect the data privacy and interests of data owners.The algorithm first builds an access control tree based on attribute combinations to ensure the thoroughness of data deletion.Then,the data deletion evidence is generated and verified using fast Fourier transform and uniqueness of polynomial coefficients to ensure the verifiability of data deletion and reduce the computational overhead using unit root.Finally,the polynomial is used to update the deleted elements to ensure the irrecoverability of data deletion.Theoretical and experimental analyses show that this algorithm can save 90% of the overhead when the number of attributes reaches 30 while ensuring data security.(2)Considering that the data owner is not always online,so in order to make it possible to delete the data stored on the remote cloud server even when the data owner is not involved,this paper proposes a data dynamic self-destruction algorithm.The algorithm converts the service duration of the data into the number of accesses and manages the access time through a binary tree to achieve fine-grained control of the data deletion time.In addition,the KUNodes technique is used to implement the update of shared data and private keys to ensure the availability of data during deletion operations.Finally,the algorithm uses a combination of binomial tree and queueing techniques to achieve dynamic update of deletion time and improve the adjustability of data self-destruction operations.Theoretical and experimental analyses show that the proposed algorithm ensures the security of data self-destruction operations with a small overhead.