根据实体自身的服务特性,从实体服务的信任关系中抽象出具有代表性的多服务属性,提出了一种基于多服务属性的信任评估模型。服务请求者根据自身喜好对服务提供者提供的多服务属性进行综合评判,并结合其信任值决定是否进行交易。交易结束后,服务请求者根据实际得到的服务质量和服务提供者自身声称的服务质量计算服务质量差异度,以此判断服务提供者的可信程度,并进行相应的奖惩和信任度更新。仿真实验表明,该模型能更加精确地评估实体的信任度,可有效地遏制各类恶意实体的攻击。当系统中恶意实体比例较高时,仍能保持较高的成功交易率。
According to the node self-service features,representative multi-service properties are abstracted and a trust evaluate model based on multi-service properties is proposed.The service requester will judge the multi-service properties of the services providers according its own preference,and decide whether the transaction will begin.After the transaction,the service requester calculates the differences in degree of quality of services based on the actual services quality received and the services quality which the service provider claimes.So as to judge the credibility of service providers,and the corresponding rewards and punishments and trust update.Simulation results show that the model can be more accurate assessment of the nodes trust can effectively curb various types of malicious node attacks.When a high malicious proportion of nodes happen,the system can still maintain a relatively high rate of successful transactions.