Cloud Computing (CC) is an Internet-centered, advanced technology that is rapidly growing. It offers numerous options for resource sharing, storage, and several features. Security concerns have emerged as one of the most significant difficulties faced by the cloud environment in recent years. Using He-Recurrent Neural Network (He-RNN) and Secret Key Double-Elliptic Curve Cryptography (SKD-ECC) with user authorization, an effective model for user identification and data security was proposed in this paper to address such problems. The training phase and the secure data transmission phase were the two vital phases in the proposed work. The following steps - data owner registration, IP spoofing, User profile database, and user identification phase occur in the training phase. The He-RNN network was trained centered on the registration details. Using Double Public Key- Digital Signature Algorithm (DOUPK-DSA), the signatures are engendered during the secure data transmission phase. Using SKD-ECC, the data encryption is executed along with spoofing the data owner’s IP address. The data’s authorization is verified with the altered data identification phase before the data is transmitted to the user. Moreover, the signature matching process verifies the data user’s authentication. The data is sent to the data user if the signature matches; otherwise, the data transfer is rejected. And the proposed method’s outcomes are analogized with prevailing algorithms. The experimental outcomes exhibited that the proposed framework outperformed existing methodologies regarding accuracy, sensitivity, and specificity, and executed the privacy-preserving process more accurately

Recurrent neural network, He initialization, Elliptic curve cryptography, Digital signature algorithm, Privacy-preserving in the cloud