ZK Hints
Daily Digest 每日摘要 View All 查看全部
01
Fri 五
Yanpei Guo PCS Error-correcting Code
Guo et al. proposed LigeSIS, the first distribution-friendly code-based multilinear PCS, achieving sublinear cross-node communication with proof size independent of machine count. Guo等人在论文中提出了首个分布友好的基于纠错码的多线性多项式承诺方案LigeSIS，通过同态子集和哈希实现亚线性跨节点通信，证明大小与机器数量无关。
Key Points: 要点：
- • LigeSIS is the first distribution-friendly code-based multilinear PCS LigeSIS是首个分布友好的基于纠错码的多线性多项式承诺方案
- • Replaces Merkle-tree hashing with homomorphic subset-sum hash over Goldilocks64 用Goldilocks64上的同态子集和哈希替代Merkle树哈希，实现代数聚合
- • Introduces preprocessing-accelerated subset-sum hash to reduce overhead 引入预处理加速的子集和哈希，降低哈希开销
- • Single-node performance comparable to state-of-the-art RS-based PCS WHIR 单节点性能媲美最先进的RS-based PCS WHIR
- • Near-linear scalability in prover time in distributed settings 分布式环境下证明时间呈现近线性扩展性
- • Improves prover time over distributed MKZG and reduces cross-node communication vs PIP 相比分布式MKZG，证明时间提升显著；相比PIP，跨节点通信减少