Hard real-time scheduling of distance-constrained sensor tasks within a monitoring system for mobile robot guidance

This paper describes a new scheduling strategy, which can be used to schedule distance-constrained non-preemptive hard real-time tasks. Besides the ability to guarantee the satisfaction of all hard time requirements, the scheduler simultaneously enables a dynamical, e.g. priority driven, distribution of free resources. The authors motivate the proposed new scheduling strategy on the basis of a monitoring and navigation system for mobile vehicle guidance in industrial environments. This system, called MONAMOVE, is currently under development in the authors´ lab. However, this application serves only as an example; similar scheduling requirements can be found in many real-time applications, where a variable number of tasks with distance-constraints share common resources and where a distribution of processing resources, not used to satisfy the hard timing requirements, is desired