Volume 23, Issue 60, July - December, 2026

An Intelligent and Privacy- Preserving Cloud Data Storage Framework Using CLCN-DOA Learning and Efficient Quantum Homomorphic Encryption

Anvar Shathik J¹, Vishnu Sakthi D², Naveen G³, Mosses A⁴, Ramshankar N⁵, Siva Subramanian R⁶

¹Dept of Computer Science & Engineering, Anjuman Institute of Technology and Management, Belalkanda, Bhatkal, Karnataka, 581320, India
²Department of Artificial Intelligence and Data Science, Easwari Engineering College, Ramapuram, Chennai, India
³Dept. of ECE, Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, SIMATS University, Kancheepuram, Chennai-602105, India
⁴Saveetha Sshool of Engineering, Professor, Department of ECE, Saveetha Institute of Medical and Technical Sciences, Chennai, India
⁵Department of CSE, Saveetha Engineering College, Chennai, Tamilnadu, India
⁶Dept of CSE, SRM Institute of Science and Technology, Chennai, India

ABSTRACT

Cloud data storage is an essential requirement under the framework of the contemporary computing environment, particularly with regard to the growing amount of sensitive data and cyber-attacks. Nevertheless, it is currently hard to guarantee strong anomaly detection and the ability to keep the data confidential because of the complex patterns of user behavior, dynamic contexts of access, and the necessity to conduct privacy-preserving computation. To address these challenges, an advanced framework is proposed, Clifford-Latent Convolutional Network with Dhole Optimization Algorithm (CLCN-DOA) facilitated by Efficient Quantum Homomorphic Encryption (EQHE) for secure cloud data management and anomaly detection. First, pre-processing is performed on the access logs in the cloud which makes use of Adaptive Sliding Window Normalization (ASWN) to remove noise from the records and normalize the behavioral features. A Convolutional Swin-Transformer Network (CSTN) is then used to extract the hierarchical and robust features. The CLCN, in turn, combines Clifford-Steerable Convolutional Neural Networks (CSCNN) and a Latent Attention Network (LAN), leveraging the strengths of both to extract both geometric and behavioral patterns (latent) from the data. The DOA optimizes network parameters to reduce the amount of false positives and improve detection accuracy, and EQHE reduces data confidentiality when stored and processed. Extensive analyses demonstrate that the suggested framework obtains a high accuracy of 99.7%, low-latency of 68 ms, and low-storage overhead of 28.6%. These results demonstrate that the recommended system is strong, effective, and capable of safely storing cloud data and detecting anomalies in reallife settings.

Keywords: Cloud Data Storage, Clifford-Latent Convolutional Network, Dhole Optimization Algorithm, Efficient Quantum Homomorphic Encryption, Convolutional Swin-Transformer Network

Indian Journal of Engineering, 2026, 23(60), e15ije1735

PDF

DOI: https://doi.org/10.54905/disssi.v23i60.e15ije1735

Published: 30 September 2026