A Hybrid Approach to Data Classification in Cloud Storage: Leveraging AES and Runge-Kutta for Optimal Security

Abstract

Secure data classification is crucial in hybrid cloud computing environments, where sensitive data may be transmitted across both private and public cloud infrastructures. This paper introduces a novel approach for secure data classification tailored to hybrid cloud settings, aiming to protect sensitive information while harnessing the advantages of hybrid architectures. The proposed method integrates cryptographic techniques, specifically AES encryption combined with the Runge-Kutta algorithm, to classify data according to its sensitivity and implement appropriate security measures. The approach entails encrypting sensitive data prior to transmission, applying fine-grained access controls that restrict data access based on classification levels, and utilizing machine learning models for automated data classification. Experimental results indicate that the AES + Runge-Kutta method achieves an encryption time of 231.21 ms and a decryption time of 219.87 ms for a 10MB file. Evaluations demonstrate the approach's effectiveness and efficiency in safeguarding sensitive data within hybrid cloud environments while minimizing performance overhead.