papers/zoo-bridge/sections/04-warp-bridge-protocol.tex

\section{Warp Bridge Protocol}

\label{sec:warp}

\subsection{Message Format}

\begin{lstlisting}[basicstyle=\ttfamily\small,caption=Zoo Bridge Warp Message]
{
  "version": 1,
  "sourceChainId": 200200,
  "destChainId": 96369,
  "nonce": 42,
  "sender": "0x<bridge_contract_zoo>",
  "recipient": "0x<bridge_contract_lux>",
  "assetType": "TOKEN|MODEL_NFT|DATA_COMMIT|CREDITS",
  "payload": "<asset-specific-data>",
  "timestamp": 1740000000,
  "aggregateSignature": "0x<bls_sig>"
}
\end{lstlisting}

\subsection{Verification}

The destination chain verifies the Warp message by:

\begin{enumerate}
    \item Fetching the source chain's validator set from the P-Chain
    \item Verifying the aggregate BLS signature against the validator public keys
    \item Checking that the signing stake exceeds the 67\% quorum threshold
    \item Verifying the nonce to prevent replay attacks
    \item Executing the bridge action (mint/unlock/create)
\end{enumerate}

\begin{theorem}[Warp Bridge Safety]
If the source chain's validator set is Byzantine fault tolerant (at most $f < n/3$ Byzantine validators), then no invalid Warp message can be verified on the destination chain.
\end{theorem}

\begin{proof}
A valid aggregate signature requires $\geq 67\%$ of validator stake. With $f < n/3$ Byzantine validators controlling $< 33\%$ of stake, they cannot produce a valid aggregate signature for a message not endorsed by honest validators. Honest validators only sign messages corresponding to finalized transactions on the source chain.
\end{proof}

\subsection{Latency Analysis}

\begin{table}[t]
\centering
\begin{tabular}{lc}
\toprule
\textbf{Phase} & \textbf{Latency (ms)} \\
\midrule
Source chain finality (Quasar) & 10 \\
Warp message construction & 5 \\
BLS signature aggregation & 80 \\
Message propagation (source $\to$ dest) & 120 \\
Destination chain inclusion & 10 \\
Signature verification & 15 \\
Bridge action execution & 8 \\
Destination finality & 10 \\
\midrule
\textbf{Total (Lux $\leftrightarrow$ Zoo)} & \textbf{258} \\
\textbf{Total with safety margin (2$\times$)} & \textbf{$\sim$500} \\
\textbf{Observed median} & \textbf{1,200} \\
\bottomrule
\end{tabular}
\caption{Warp bridge latency breakdown. Observed median is higher due to validator coordination overhead.}
\label{tab:warp-latency}
\end{table}

The gap between theoretical (258ms) and observed (1.2s) latency is due to validator coordination: collecting sufficient BLS partial signatures requires waiting for slow validators. We use a 67\% stake threshold, so the bridge waits until enough validators respond.