mirror of
https://github.com/lux-apps/papers.git
synced 2026-10-30 18:15:21 +00:00
Restructured from flat .tex/.pdf files into per-paper subdirectories. Added Antje Worring as co-author on all papers. Synced PDFs to pdfs/ for download site.
47 lines
1.9 KiB
TeX
47 lines
1.9 KiB
TeX
\section{Validator Selection and Abstraction}
|
|
\label{sec:validators}
|
|
|
|
\subsection{Validator Set Management}
|
|
|
|
The StakingPool delegates to a curated set of validators selected by governance:
|
|
|
|
\begin{algorithm}[H]
|
|
\caption{Validator Selection and Rebalancing}
|
|
\label{alg:validator}
|
|
\begin{algorithmic}[1]
|
|
\Require Candidate validators $\mathcal{V}_{\text{candidates}}$, selection criteria
|
|
\State $\mathcal{V}_{\text{active}} \gets$ top $K$ validators by score
|
|
\State Score $= w_u \cdot \text{uptime} + w_c \cdot (1 - \text{commission}) + w_d \cdot \text{decentralization}$
|
|
\For{each epoch}
|
|
\State Compute target delegation per validator: $\delta_i = \mathsf{totalPooled} / |\mathcal{V}_{\text{active}}|$
|
|
\For{each validator $v_i \in \mathcal{V}_{\text{active}}$}
|
|
\State $\Delta_i \gets \delta_i - \text{currentDelegation}(v_i)$
|
|
\If{$|\Delta_i| > \mathsf{rebalanceThreshold}$}
|
|
\State Delegate or undelegate $\Delta_i$ LUX
|
|
\EndIf
|
|
\EndFor
|
|
\EndFor
|
|
\end{algorithmic}
|
|
\end{algorithm}
|
|
|
|
\subsection{Decentralization Constraints}
|
|
|
|
\begin{definition}[Delegation Cap]
|
|
No single validator receives more than $\mathsf{maxDelegationShare} = 10\%$ of the total staking pool:
|
|
\begin{equation}
|
|
\forall\, v_i : \text{delegation}(v_i) \leq 0.10 \cdot \mathsf{totalPooled}
|
|
\end{equation}
|
|
\end{definition}
|
|
|
|
With a minimum of 10 validators, this ensures that compromise of any single validator affects at most 10\% of the pool.
|
|
|
|
\begin{proposition}[Decentralization Preservation]
|
|
If the LLS staking pool holds $P$ of total stake $S$, the Nakamoto coefficient satisfies:
|
|
\begin{equation}
|
|
\mathsf{NC} \geq \left\lceil \frac{S/3}{0.10 \cdot P + \max_{j \notin \text{LLS}} s_j} \right\rceil
|
|
\end{equation}
|
|
\end{proposition}
|
|
|
|
For $P = 0.3S$ (LLS holds 30\% of stake): an attacker needs the entire LLS pool plus additional stake, requiring compromise of governance (multi-sig + timelock) rather than a single validator.
|
|
|