Skip to content

Transciphering Framework for Approximate Homomorphic Encryption

Authors: Cho, Ha, Kim, Lee, Lee, Lee, Moon, Yoon; Ha, Kim, Lee, Lee, Son
Year: 2021
Venue: ASIACRYPT 2021; EUROCRYPT 2022

type: scheme collection: schemes status: auto

Links: ePrint 2020/1335

Visual lesson: HERA / Rubato (CKKS transciphering) · Scheme table row: HERA / Rubato (CKKS transciphering)

Summary

Auto-extracted summary (unreviewed)

Introduces the RtF (Real-to-Finite-field) transciphering framework, which chains the FV (BFV) and CKKS schemes with a modular-arithmetic stream cipher so that clients can encrypt real-valued data with symmetric-cipher cost and small ciphertext expansion while the server converts the result into CKKS ciphertexts. The authors also design HERA, a new HE-friendly stream cipher with a simple randomized key schedule that needs fewer random bits than FLIP or Rasta. At 128-bit security, RtF with HERA achieves a ciphertext expansion ratio of 1.23-1.54 (at least 23x smaller than CKKS-only) and client-side latency and throughput of 1.6 us and 21.7 MB/s, 9085x and 17.8x better than CKKS-only.

Abstract

Homomorphic encryption (HE) is a promising cryptographic primitive that enables computation over encrypted data, with a variety of applications including medical, genomic, and financial tasks. In Asiacrypt 2017, Cheon et al. proposed the CKKS scheme to efficiently support approximate computation over encrypted data of real numbers. HE schemes including CKKS, nevertheless, still suffer from slow encryption speed and large ciphertext expansion compared to symmetric cryptography. In this paper, we propose a novel hybrid framework, dubbed RtF (Real-to-Finite-field) framework, that supports CKKS. The main idea behind this construction is to combine the CKKS and the FV homomorphic encryption schemes, and use a stream cipher using modular arithmetic in between. As a result, real numbers can be encrypted without significant ciphertext expansion or computational overload on the client side. As an instantiation of the stream cipher in our framework, we propose a new HE-friendly cipher, dubbed HERA, and extensively analyze its security and efficiency. The main feature of HERA is that it uses a simple randomized key schedule. Compared to recent HE-friendly ciphers such as FLIP and Rasta using randomized linear layers, HERA requires a smaller number of random bits. For this reason, HERA significantly outperforms existing HE-friendly ciphers on both the client and the server sides. With the RtF transciphering framework combined with HERA at the 128-bit security level, we achieve small ciphertext expansion ratio with a range of 1.23 to 1.54, which is at least 23 times smaller than using (symmetric) CKKS-only, assuming the same precision bits and the same level of ciphertexts at the end of the framework. We also achieve 1.6 µs and 21.7 MB/s for latency and throughput on the client side, which are 9085 times and 17.8 times faster than the CKKS-only environment, respectively. ⋆This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No.2021R1F1A1047146).
Source: pdf

Extracted details

Field Value Source Confidence
contribution RtF (Real-to-Finite-field) transciphering framework combining FV and CKKS with a modular stream cipher llm 0.85
contribution HERA, a new HE-friendly stream cipher with a randomized key schedule llm 0.85
contribution security analysis of HERA against algebraic and other attacks llm 0.85
contribution implementation with client-side encryption in C++ and server-side transciphering in Lattigo, compared against PEGASUS and CKKS-only llm 0.85
contribution comparison of HERA with existing HE-friendly ciphers using HElib llm 0.85
schemes CKKS llm 0.85
schemes BFV llm 0.85
schemes HERA stream cipher llm 0.85
libraries Lattigo llm 0.85
libraries HElib llm 0.85
tasks transciphering / hybrid homomorphic encryption llm 0.85
tasks efficient client-side encryption of real-valued data for CKKS llm 0.85
security level 128-bit llm 0.85
threat model client-server (client encrypts with a symmetric cipher, server transciphers to CKKS) llm 0.85
interaction non-interactive FHE llm 0.85
code url https://github.com/KAIST-CryptLab/RtF-Transciphering (URL found in the full text) regex 0.50
code url https://github.com/LowMC/lowmc/blob/master/determine_rounds.py (URL found in the full text) regex 0.50
code url https://github.com/Alibaba-Gemini-Lab/OpenPEGASUS (URL found in the full text) regex 0.50
code url https://github.com/KAIST-CryptLab/RtF-Transciphering llm 0.85
llm relevance low llm 0.85
key result ciphertext expansion ratio: 1.23-1.54 (at least 23x smaller than CKKS-only) (RtF + HERA, 128-bit security, same precision and final ciphertext level) llm 0.85
key result client-side encryption latency: 1.6 us (9085x faster than CKKS-only) (RtF + HERA client side) llm 0.85
key result client-side encryption throughput: 21.7 MB/s (17.8x faster than CKKS-only) (RtF + HERA client side) llm 0.85

Scheme card: HERA / Rubato (CKKS transciphering)

Group 7. Scheme switching and hybrid (transciphering) HE
Kind Scheme
Hardness assumption Symmetric cipher + RLWE
Plaintext type Approximate real (to CKKS)
SIMD packing Yes
Bootstrapping Uses CKKS bootstrapping
Post-quantum Partly (cipher-dependent)
Status Active
Fit for LLM inference Client sends small symmetric ciphertexts and the server converts them to CKKS. This cuts upload size for long prompts.
Follow-ups Rubato (2022/537); key-recovery attack on Rubato (2023/822); HHE survey (2025/071).
Follow-up papers in the archive Rubato : Noisy Ciphers for Approximate Homomorphic Encryption (Full Version)
Cryptanalysis of Symmetric Primitives over Rings and a Key Recovery Attack on Rubato
The HHE Land: Exploring the Landscape of Hybrid Homomorphic Encryption

Keyword mentions

Counts of vocabulary terms in the full text (first page of occurrence in brackets).

  • scheme: CKKS 74 [p.1], Hybrid HE / transciphering 35 [p.1], MPC 6 [p.18], BFV 5 [p.4], BGV 5 [p.3], TFHE 3 [p.6], Scheme switching 3 [p.5]
  • library: HElib 4 [p.5], Lattigo 4 [p.5]
  • model: SVM 1 [p.30]
  • technique: Bootstrapping 20 [p.5], Rescaling 9 [p.8], Polynomial approximation 5 [p.9], Rotations 4 [p.19], Chebyshev 3 [p.9], Minimax approximation 2 [p.28], NTT 2 [p.9], RNS 2 [p.24]
  • assumption: LWE 14 [p.6], RLWE 12 [p.6], Lattice-based 3 [p.29], 128-bit security 3 [p.1]
  • task: Medical / healthcare 3 [p.1], Finance 3 [p.28], Language modeling / generation 2 [p.9]
  • topic: Homomorphic encryption 61 [p.1], Open source 8 [p.5], Benchmark 3 [p.25]

Files

Path (under papers/) Collection Pages Size SHA-256 Note
references/schemes/s41-0_transciphering-framework-for-approximate-homomorphic-encrypt.pdf schemes 46 543.3 KB 9f24707e0c6b

Database key: transciphering-framework-for-approximate-homomorphic-encryption · created 2026-11-05 · updated 2026-11-05 · edit with python3 tools/paperdb.py set transciphering-framework-for-approximate-homomorphic-encryption COLUMN VALUE