Advances and Challenges in Fully Homomorphic Encryption: Performance Optimization, Security, and Quantum-Resilient Architectures

Fully Homomorphic Encryption (FHE) has emerged as a foundational pillar in contemporary cryptography, enabling direct computations on encrypted data while preserving privacy. Despite its considerable promise across diverse sectors, such as healthcare, finance, cloud computing, and the Internet of Things, practical deployment has been hindered by high computational costs, primarily attributed to bootstrapping and noise-reduction processes. To overcome these challenges, recent research efforts have integrated algorithmic optimizations with hardware acceleration techniques, including GPUs, FPGAs, advanced caching, and purpose-built architectural designs, specifically targeting latency and scalability bottlenecks. Prominent systems, such as F۱, BTS, and Cheetah, utilize optimized polynomial arithmetic and Number Theoretic Transform (NTT) acceleration, resulting in a more than hundredfold increase in bootstrapping efficiency through the employment of memory-centric and parallel processing techniques. Concurrently, the emergence of quantum computing has redirected attention to lattice-based, quantum-resistant cryptographic frameworks, which are grounded in the hardness assumptions of Learning with Errors and Ring-LWE. Furthermore, hybrid architectures integrating hierarchical memory management and threshold-oriented quantum FHE schemes are being proposed to offer resilience against post-quantum adversaries, thereby facilitating distributed and privacy-preserving analytical capabilities. This survey synthesizes current advances across software, hardware, and quantum domains, critically evaluating their performance metrics and resistance strategies. The analysis highlights emerging design patterns that synthesize efficiency, scalability, and quantum security, and quantum safety—positioning FHE as a deployable primitive for secure computation in real-world environments.

کلیدواژه ها:

Fully Homomorphic Encryption (FHE) ، GPU Acceleration ، Post-Quantum Security ، Bootstrapping Optimization ، Architecture-Centric Design