学习驱动的自适应审计调度:一种用于链下数据完整性的序贯决策方法
Learning-Driven Adaptive Audit Scheduling: A Sequential Decision Approach to Off-Chain Data Integrity
浏览论文内容
中文总结 AI 辅助
研究链下数据完整性审计问题,提出DRQN-CMDP方法,结合GRU层与拉格朗日对偶上升,利用无配对同态MAC原语,相比其他13种方法,在燃气成本、误报率和检测延迟上取得良好平衡。
中文摘要 AI 辅助
我们将链下数据的加密审计建模为部分可观测性下的约束马尔可夫决策过程(CMDP):存储节点的隐藏类型和损坏状态使问题成为部分可观测马尔可夫决策过程(POMDP),而误报率上限ρ施加了明确的安全约束。我们提出了DRQN-CMDP,一种深度循环Q网络,其门控循环单元(GRU)层维持对潜在节点类型的信念,并与拉格朗日对偶上升相结合,自动调整误报率惩罚λ。一种无配对的同态消息认证码(MAC)原语提供O(1)的链上验证成本。在13种方法中,DRQN-CMDP实现了良好的平衡:比固定高频审计的燃气成本低83%,误报率为个位数(7.5%),检测延迟适中,这是其他方法在所有三个目标上无法同时匹配的组合。
英文摘要
We model cryptographic auditing of off-chain data as a Constrained MDP (CMDP) under partial observability: the storage node's hidden type and corruption state make the problem a POMDP, while a miss-rate ceiling rho imposes an explicit security constraint. We propose DRQN-CMDP, a Deep Recurrent Q-Network whose GRU layer maintains a belief over the latent node type, paired with Lagrangian dual ascent that adapts the miss-rate penalty lambda automatically. A pairing-free homomorphic-MAC primitive supplies O(1) on-chain verification cost. Across 13 methods--four DQN variants, PPO, A2C, PPO-Lagrangian, a stateful Bayesian heuristic, three fixed-rule baselines, and an oracle-informed heuristic--DRQN-CMDP achieves a favourable balance: 83% lower gas than fixed high-frequency auditing, single-digit miss rate (7.5%), and moderate detection latency--a combination no other method matches across all three objectives simultaneously.
发表机构
- Hangzhou Yunphant Network Technology Co., Ltd.(杭州云幻网络科技有限公司)
- Zhejiang University(浙江大学)
- Central University of Finance and Economics(中央财经大学)
- University of North Georgia(北佐治亚大学)
- Zhejiang Shuren University(浙江树人大学)
机构由 AI 辅助整理,请以论文原文为准。