Increasing the Robustness of Self-Organizing Systems By Means of Trust Practices

Self-organizing systems are becoming increasingly complex in their organisational structures, especially when unknown heterogeneous entities might arbitrarily enter and leave the network at any time. Therefore, new ways have to be found to develop and manage them. One way to overcome this issue is trust. Using appropriate trust mechanisms, entities in the system can have an indication about which entities to cooperate with. This is very important to improve the robustness of self-organizing systems, which depends on a cooperation of autonomous entities. The contributions of this thesis are trustworthy concepts and generic self-* properties that work in a distributed manner and with no central control to ensure robustness. The self-* properties examined in this thesis are self-configuration, selfoptimization, and self-healing. We believe that these properties are fundamental for the design of every autonomous, scalable and fault-tolerant self-* middleware. The self-configuration is related to the ability to perform an initial distribution of services on nodes taking the resource requirement and importance level of services into account. The self-optimization focuses on optimizing the allocation of services at runtime by monitoring the trust and resource consumption of nodes. And the self-healing aspect is concerned with the ability to handle unexpected disturbances in the system in order to guarantee that all services running on nodes stay available even in case of crash, execution and reachability failures. The resulting middleware is TEM, a trust-enabling middleware that can profit from the advantage of trust and self-* properties at the same time. The TEM makes use of different trust metrics, i.e., such as direct trust, reputation, and confidence to monitor the behavior of nodes in the system. These techniques have been used to improve the robustness of the self-* properties, both at design time and at runtime. Moreover, we applied the TEM middleware to different application case studies and clarified how uncertainty in open environments can be mastered by using the approaches investigated in this thesis. Due to the fact that future application services will become more autonomous, we expect to see more self-organizing systems based on our (or similar) approaches. The investigations conducted within this thesis are a step in this direction.