Skip to content
Open access

FHE-Lab: A Modular Framework for Customizable Fully Homomorphic Encryption Pipelines

Sep 2026 · IEEE International Symposium on Workload Characterization · pp. 444-455 · 0 citations · 35 references

Abstract

Fully Homomorphic Encryption (FHE) enables computation on encrypted data without requiring decryption, making it a foundational tool for privacy-preserving computation in untrusted environments. Modern open-source FHE libraries such as OpenFHE and HEaaN support multiple schemes and highly optimized implementations, but their layered abstractions can limit visibility into low-level operations. In particular, critical optimizations are often tightly integrated, making them difficult to independently configure or isolate for experimental analysis.To address these limitations, we present FHE-Lab, a modular FHE research framework designed to prioritize configurability, instrumentation, and observability throughout the FHE pipeline. FHE-Lab currently targets CKKS, a widely used scheme for approximate homomorphic computation, and enables researchers to customize pipeline configurations, independently control optimizations such as RNS and NTT, and inspect intermediate computation states. We demonstrate these capabilities through a case study of CKKS reliability under transient hardware errors, evaluating how different optimization and parameter configurations affect error behavior. By providing fine-grained control over the FHE pipeline, FHE-Lab aims to enable researchers to more easily explore new research questions and implementation strategies in FHE.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.