Application Level Fault Diagnosis in Organic Computing Using Stress Hormone Propagation

In the domain of Organic Computing, the utilization of biological principles in technical systems has led to the development of decentralized mechanisms like the Artificial Hormone System (AHS). In distributed real-time embedded systems, the AHS serves as middleware, continuously monitoring and organizing task allocations among computing nodes. By introducing different types of artificial hormones for the tasks, task allocations are realized by constantly establishing hormone balances via distributed closed control loops. This process handles the increasing complexity of distributed control systems by enabling self-configuration, -adaptation, -improvement and - healing. However, the occurrence of faults within the artificial hormone system, such as incorrect hormone values, can lead to significant consequences, including adverse system behavior and potentially complete system failure. To mitigate these risks, this paper proposes the integration of stress hormones within the AHS framework, focusing on fault diagnosis at the application level. Utilizing example scenarios generated by the Stanford Network Analysis Platform (SNAP), the efficacy of the stress hormone algorithm has been evaluated. These preliminary findings suggest promising opportunities for improving system resilience and robustness through the integration of stress hormones within the AHS framework.