Non-preemptive Scheduling of Distance Constrained Tasks Subject to Minimizing Processor Load

Scheduling problems that involve distance constrained tasks with repetitive requests are enjoying an increase in popularity in real time system design. Timing constraints defined by periods of fixed lengths for each task enforce some frequency of task execution. Alternatively deadlines could be specified relatively to the finish time of the previous execution of the same task. Hence, the scheduling algorithms deal with tasks that virtually consist of infinite chains of jobs with specified temporal distance constraints between each pair of adjacent jobs. In this paper, unlike approaches that consider a preemptive distance constrained task scheduling subject to minimize time distance between task execution, we consider a problem of non preemptive tasks systems with objective to reduce processor utilization while keeping the response times within their limits. As a consequence, free processor capacity remains to process aperiodic tasks. Moreover, as a side effect, the way the tasks get scheduled to obtain mnimal processor utilization results in jitter reduction improving system stability.