mirror of
https://github.com/lux-apps/papers.git
synced 2026-11-02 11:39:52 +00:00
76 lines
2.1 KiB
TeX
76 lines
2.1 KiB
TeX
\section{Comparison with Alternative PQC Schemes}
|
|
|
|
|
|
\subsection{ML-DSA}
|
|
|
|
ML-DSA \cite{ducas2018crystals} is NIST's primary lattice-based signature standard:
|
|
|
|
\textbf{Advantages:}
|
|
\begin{itemize}
|
|
\item Simpler implementation (no FFT sampling)
|
|
\item Deterministic signing (no randomness needed)
|
|
\item Slightly faster verification
|
|
\end{itemize}
|
|
|
|
\textbf{Disadvantages:}
|
|
\begin{itemize}
|
|
\item Larger signatures: Dilithium2 = 2420 bytes (3.6x FALCON-512)
|
|
\item Larger public keys: 1312 bytes vs. 897 bytes
|
|
\item Higher bandwidth cost
|
|
\end{itemize}
|
|
|
|
\textbf{EVM Feasibility:}
|
|
\begin{itemize}
|
|
\item Estimated 3-4M gas (2x ETHFALCON)
|
|
\item NTT also required (same optimization applies)
|
|
\item No clear advantage for Ethereum deployment
|
|
\end{itemize}
|
|
|
|
\subsection{SLH-DSA (FIPS 205, formerly SPHINCS+)}
|
|
|
|
SLH-DSA \cite{bernstein2019sphincs} is a hash-based signature scheme:
|
|
|
|
\textbf{Advantages:}
|
|
\begin{itemize}
|
|
\item Provably secure (minimal assumptions)
|
|
\item No structured assumptions (no lattices)
|
|
\item Simple implementation
|
|
\end{itemize}
|
|
|
|
\textbf{Disadvantages:}
|
|
\begin{itemize}
|
|
\item Massive signatures: SLH-DSA-128f = 17,088 bytes
|
|
\item Slow signing: $>$100ms for fast variant
|
|
\item Calldata cost: 68k gas (vs. 2.6k for FALCON)
|
|
\end{itemize}
|
|
|
|
\textbf{EVM Feasibility:}
|
|
\begin{itemize}
|
|
\item Estimated 5-10M gas (verification cost)
|
|
\item Calldata alone costs 1.1M gas
|
|
\item Impractical for Ethereum deployment
|
|
\end{itemize}
|
|
|
|
\subsection{Summary Comparison}
|
|
|
|
\begin{center}
|
|
\small
|
|
\begin{tabular}{lrrrr}
|
|
\hline
|
|
\textbf{Scheme} & \textbf{PubKey} & \textbf{Sig} & \textbf{EVM Gas} & \textbf{Status} \\
|
|
\hline
|
|
ECDSA & 33 B & 65 B & 3k & Current \\
|
|
FALCON-512 & 897 B & 666 B & 7M & This work \\
|
|
ETHFALCON & 897 B & 666 B & 1.8M & This work \\
|
|
EPERVIER & 20 B* & 2202 B & 1.9M & This work \\
|
|
Dilithium2 & 1312 B & 2420 B & 3-4M & Estimated \\
|
|
SLH-DSA-128f & 32 B & 17088 B & 5-10M & Estimated \\
|
|
\hline
|
|
\end{tabular}
|
|
\end{center}
|
|
|
|
* Address only (hash of public key)
|
|
|
|
\textbf{Conclusion:} ETHFALCON/EPERVIER offer the best balance of security, performance, and Ethereum compatibility among NIST PQC candidates.
|
|
|