Author(s)

Nan Hu, Zhenjiang Lei, Lei Wang and Yubo Liu

Corresponding Author

Nan Hu

Abstract

Identity authentication scheme that combines security, recognition rate, and operational efficiency in cloud computing environment is proposed in this paper, which utilizes fully homomorphic encryption on integer ring optimized to protect identity feature information security, and neural network model is used to identity feature extraction. Moreover, by converting Euclidean distance calculation among identity feature vectors into dot product calculation during identity recognition, computational complexity is reduced. Finally, based on Store Front authentication architecture of Citrix desktop cloud system, identity-based desktop cloud authentication system is designed, and identity authentication scheme is implemented through API interface provided by Citrix Store Front component. What’s more, experimental verification is performed in actual distributed environment so that security and feasibility of solution can be analyzed, which shows that this solution can not only improve security of identity authentication, but also have high recognition rate and operating performance to meet requirements of practical applications.

Keywords

Cloud computing; Neural network model; Identity authentication framework