Multi-objective reconfiguration of evolving real-time microservices-based environments using energy-based nuclear reaction optimization
Maryam Nooraei Abadeh · Systems and Soft Computing · 2025
In modern systems, which are increasingly dynamic and in continuous need of adaptation, one of the main challenges involves ensuring resilience during the management of evolving architecture. This paper describes the application of Nuclear Reaction Optimization in the reconfiguration of evolving real-time microservices based environments to enhance resilience through automatic repair mechanisms. Guided by the respective fusion and fission processes, NRO merges (fusion) components to improve fault tolerance and splits them (fission) in order to improve modularity, dynamically adapting the system. Simulation and case studies have demonstrated that the NRO-based approach effectively balances performance, resource utilization, and fault tolerance, outperforming Particle Swarm Optimization (PSO) and "Non-dominated Sorting Genetic Algorithm II (NSGA-II) in resilience and resource management. The following study, therefore, illustrates the potential of NRO within the optimization process of multi-objective complex adaptive systems to support the resilience and evolution process of real-world applications.