DIAGNOSIS TASKS SCHEDULING BASED ON A DISTRIBUTED MULTIPROCESSORS SYSTEM

This paper presents study results of real-time tasks scheduling of distributed diagnosis by genetic algorithm. Diagnoses concern distributed facilities and are performed on real-time resources. The problem belongs to the category of job-shops, with shared, distributed and networked resources. Resources are assigned to tasks led by algorithms, also considered as mobile agents. The proposed scheduling solution based on GA enables to minimize the sum of makespan on the whole of the diagnosis processors, while maximizing the availability of the distributed facilities.