A REAL-TIME SCHEDULER USING NEURAL NETWORKS FOR SCHEDULING INDEPENDENT AND NONPREEMPTABLE TASKS WITH DEADLINES AND RESOURCE REQUIREMENTS

This paper describes a neural network scheduler for scheduling indepen dent and nonpreemptable tasks with deadlines and resource requirements in critical real-time applications, in which a schedule is to be obta ined within a short time span. The proposed neural network scheduler i s an integrate model of two Hopfield-Tank neural network models. To co pe with deadlines, a heuristic policy which is modified from the earli est deadline policy is embodied into the proposed model. Computer simu lations show that the proposed neural network scheduler has a promisin g performance, with regard to the probability of generating a feasible schedule, compared with a scheduler that executes a conventional algo rithm performing the earliest deadline policy.