安全多方计算中一般合作计算者之间互不信任,为保护私有信息通常采用零信息泄漏的安全策略,从而增加了合作计算复杂性和通信复杂性。提出了一种基于反馈的信任生成算法,通过参与者的历史行为评估其信任度,将参与者的信任度量化表示。根据参与者的信任度采取相应级别的安全措施,可降低合作的计算复杂性和通信复杂性,隔离信任值低的恶意参与者。仿真实验结果表明,该算法在安全多方计算环境中可有效遏制恶意参与者,提高合作成功率。
In secure multi-party computation,all cooperation participants are usually regarded distrust,in order to protect private information usually use zero information leakage security policy.This security policy greatly increases computational complexity and communication complexity.Propose a trust generating algorithm,through participants' history behaviors to assess their trust,express trust value in a quantitative way.Taking security policy according to participants' trust value can greatly reduce the cooperation computing complexity and insulate malicious participants.Experimental results show that this algorithm can effectively be against malicious participants and increase the ratio of success cooperation.