Partitioned Scheduling of Implicit-deadline Sporadic Task Systems under Multiple Resource Constraints

On many multiprocessor platforms each individual processor may have limited amounts of several different kinds of resources such as computing capacity, local memory, and network bandwidth. In order to partition tasks effectively upon such platforms the partitioning algorithm should be cognizant of all the resource constraints. We present and evaluate an algorithm for partitioning a collection of implicit-deadline sporadic tasks upon a multiprocessor platform in a manner that is cognizant of multiple such resource constraints.