Research On Software-Defined Cloud Storage For User Data Security

With the arrival of the big data era, users are more and more attention to the data, and the traditional storage cannot meet the explosive growth of data, then cloud storage came into being. Cloud storage is a new concept derived and developed from cloud computing, and an emerging network storage technology referring to a system which employs application software to achieve and facilitate coordination between a great many different types of storage devices by means of cluster application, network technology or distributed file system to provide data storage and service access functions for users.Cloud storage provides users with an on-demand storage service, but when users hand data over to the custody of a cloud storage service provider, they have to consider the problem of data security. Data security involves three aspects: data confidentiality, data integrity and data availability. Conventionally, the assurance of data confidentiality depends on various encryption algorithms; the assurance of data integrity relies mainly on audit program disclosure, and the audit programs needing to be disclosed fall broadly into two categories: POR and PDP; for data integrity,a redundancy strategy is mostly adopted. Redundancy strategy mainly includes backup and coding. But there are some problems in the traditional solutions of data confidentiality and integrity: 1. High computational complexity is unavoidable for both encryption algorithm and public audit,so a huge computing cost will be caused. For the users that have limited computing power and need to store massive data, such a computing cost may be unaffordable; 2. The cryptography-based method leads to problem of key management, and user privacy will leak in case of key exposure;moreover, key loss will deprive users of their ability to retrieve data.In order to keep users'data safe, this paper started with architecture,proposing a storage solution based on software-defined cloud storage. This solution is based on software-defined cloud storage architecture, aimed at storing data in different cloud storage service providers that have a competitive relation with each other, to ensure data confidentiality; a file retrieval probability model based on chunked replica redundancy was set up. Storage solution was modeled in accordance with file retrieval probability and expense, and then the model was solved by genetic algorithm.This paper made practical research and statistics on the numerical values needed in the cloud storage solution, adopted relevant examples for modeling and its solution, and evaluated the correctness and feasibility of the model. The result shows that file chunking and chunked replica backup can help greatly increase the probability of data retrieval, and that this file has an obvious protective effect on file confidentiality and integrity, so this file saving solution can well ensure the security of users' data.