摘要： 区块链技术的广泛应用导致其隐私泄露问题日益严重。为有效评估区块链技术存在的隐私泄露风险，从区块链技术架构的角度，通过基于博弈的方法对攻击进行量化，提出一种区块链隐私泄露风险评估方法。考虑用户对区块链技术架构层中各攻击的敏感性程度不同，构建用户敏感度矩阵，分别计算主观敏感度与客观敏感度。利用区块链诚实用户与恶意攻击者之间的策略交互过程构建不完全信息静态贝叶斯博弈模型，通过双方的期望收益定义风险影响性与可能性，从而得到基于区块链技术架构层的风险评估分数以及整个区块链的隐私泄露风险评估分数。在此基础上，利用Sigmoid函数对风险评估分数进行正则化处理并根据风险指数判断隐私泄露的风险等级，将隐私泄露分为风险可忽略、风险适中以及风险异常3种状态。实验结果表明，该方法能够有效评估区块链隐私泄露风险情况，指导用户进行多层次、有针对性的隐私保护。

关键词: 区块链, 博弈论, 隐私泄露, 纳什均衡, 风险评估

Abstract: The widespread application of blockchain technology has led to the increasingly serious problem of privacy disclosure.To effectively assess this risk from the perspective of blockchain technology architecture, this study proposed a method for quantifying attacks using game based methods.In considering the different sensitivity of users to attacks in the blockchain technology architecture layer, a user sensitivity matrix is constructed to calculate subjective and objective sensitivity.The policy interaction process between honest users and malicious attackers on the blockchain is used to build a static Bayesian game model with incomplete information, and the risk impact and possibility are defined by the expected benefits of both parties, thereby obtaining the risk assessment score based on the blockchain technology architecture layer and privacy disclosure risk assessment score of the entire blockchain.Based on this, a Sigmoid function is used to regularize the risk assessment scores and determine the risk level of privacy leakage according to the risk index.Privacy leakage is divided into three risk states:negligible, moderate, and abnormal.The experimental results show that this method can effectively assess the risk of blockchain privacy disclosure and ensure multi-level and targeted privacy protection to users.

Key words: blockchain, game theory, privacy disclosure, Nash equilibrium, risk assessment

中图分类号: 

• TP391