A Mechanism for Minimizing Trust Conflicts in Organic Computing Systems

In a trusted organic computing system with nodes representing client machines, the trust behavior of node participants may change during runtime. Every node in the system can execute services of other nodes and also provide services to other nodes. It has to decide using trust for which nodes it wants to execute the services and to which nodes it wants to give its services. This service distribution is performed by using our former developed trust-based self-configuration and self-optimization approach, which was devised to improve the availability of important services. However, it works only well for environments, where all nodes see the same trust value for a certain node. Hence, the baseline algorithm is not suited to operate with conflicting trust values. This situation can occur by collecting trust values independently from the neighbors of a node that can contradict each other. The contribution of this work proposes a conflict resolution mechanism as an extension to the baseline algorithm to operate with conflicting trust values at the same time.