Optimal Job Scheduling in Real-Time Cyber-Physical Systems: A Soft-Computing Approach

Real-time cyber-physical systems in aerospace and other safety-critical domains require deterministic task execution under strict certification constraints. Although static time-triggered scheduling fulfills these requirements, its manual configuration remains error-prone and limits scalability. This article proposes an automated framework that generates optimized schedules for fixed-priority, fully preemptive systems based on the Thales Alenia Space—Italia real-time platform. A simulator that replicates the scheduling semantics of the target platform is implemented, enabling precise analysis and validation of a given schedule plan. The scheduling problem is formulated as a bilevel optimization solved via soft-computing techniques, while a complementary mixed-integer linear programming formulation provides performance bounds. The approach is validated on an industrial satellite case study, demonstrating a significant reduction in preemptions and improved timing consistency compared to manually engineered schedules.