面向信任的协同模型

Coordination model for trust

针对开放分布式系统中收集实体的经验值和度量信任困难的现状,采用关注分离的原则,将实体的信任处理行为与计算行为、交互行为进行分离,提出了一个面向信任的协同模型。通过模型中的信任角色实时监听所绑定的计算实体的消息来更新经验值,进而计算信任度,为系统选择相对可信的计算实体提供了依据。最后,利用Java消息服务技术和EJB技术给出了该模型的实现框架。

To solve the difficulties of collecting experience and measuring trust of entities in open and distributed systems, the paper separated the behavior of dealing with trust problems from computing and communicating behaviors of entities by adopting the principle of separation of concerns, and presented a coordination model for trust. Trust-roles of the model could compute trust degree of bundled computing-entities by monitoring their messages to update their experience in real-time, so that the model provided evidence for the system to choose relatively credible computing-entities. Finally, implemented the infrastructure of the model by using the Java message service technology and the EJB technology.