一种基于动态规划的自动信任协商策略

Dynamic programming-based strategy for automated trust negotiation

动态规划是解决多阶段决策过程最优化的一种数学方法,可将其运用到自动信任协商中。针对目前有关协商策略的研究中没有区分信任凭证的敏感度和格式的问题,引入披露开销的概念,设计了一种新的协商策略。该策略采用动态规划的思想,基于与/或图建模,分解协商过程,自底向上求解最小开销的凭证披露序列。经证明,该策略是可采纳且高效的,能保障协商的安全性和提高协商的效率。

Dynamic programming, an approach to solve the optimal question, was introduced into automated trust negotiation. The disclose cost was introduced to solve the problem of treating all credentials with the same sensitivity and the same format. A negotiation strategy modeled on AND/OR graph was proposed under the idea of dynamic programming to achieve the global optimization. The strategy achieved the minimal cost disclose sequence by traversing the AND/OR graph upward. The strategy is proved to be adoptable and efficient to ensure the safety and efficiency of the negotiation.