An Evolutionary Method To Solve The Real-Time Scheduling Problem

FERRO, E.; CAYSSIALS, R.; ALIMENTI, O.; JAVIER DARIO OROZCO

Palma de Mallorca

Congreso; ASC 2006; 2006

In this paper we proposed a method to solve the problem of assigning and scheduling a set of preemptible periodical real-time control tasks on a set of heterogeneous processors. It is based on a stochastic optimization of a genetic-heuristic algorithm. The tasks are interrelated by both mutual exclusion and precedence constraint on multiple processors. The system is considered to meet all hard real-time constraints, such as: preallocation, precedence, memory and communication constraints. A control system example is presented.