Optimizing Secure Cloud Computation: A Comparative Study of Fully Homomorphic Re-encryption Models for Encrypted Data Evaluation

Abstract

Fully Homomorphic Encryption (FHE) enables com- putation on encrypted data without decryption, offering a promising solution for secure cloud computing. However, noise accumulation during homomorphic operations necessitates re- encryption (bootstrapping) to maintain computational correct- ness. This paper presents a comprehensive comparative study of FHE re-encryption models, analyzing their performance, security guarantees, and practical applicability in cloud environments. We evaluate multiple re-encryption schemes across key metrics including computational overhead, ciphertext expansion, noise management efficiency, and scalability. Our experimental analysis demonstrates trade-offs between security levels and computa- tional efficiency, providing insights for selecting appropriate FHE re-encryption models based on application requirements. The findings reveal that optimized bootstrapping techniques can reduce re-encryption overhead by up to 40% while maintaining equivalent security levels, making FHE more practical for real- world cloud computation scenarios.