Improved hybrid scheduling design of partitioned processing and network communication in distributed integrated modular avionics systems

Traditional hybrid scheduling method of partitioned processing and network communication restricts the schedulability of distributed integrated modular avionics (DIMA) systems according to strict constraints. This paper loosens these constraints by introducing harmonic partitions, converts the hybrid scheduling design into a nonlinear optimization problem, and proposes an improved method based on the tabu search metaheuristic. Comparative experiments demonstrate that this method adapts to a larger-scale system including 50 partitions and 1000 messages and guarantees lower processing-level end-to-end delay with an average optimization of 48%.