An Approach to Task Allocation for Dynamic Scheduling in Reconfigurable Computing Systems

Abstract

Recent research has expanded to encompass a view of the reconfigurable logic where a resource is shared among multiple applications or users. Runtime reconfiguration is a method of computing by dynamically changing hardware configurations from phase to phase of a computation. In real-time multitasking system design, task deadlines play an important role as systems not only need to support sharing of the resources in space, but also need to guarantee execution of the tasks. At the operating system level, sharing logic gates, wires, and I/O pins among multiple tasks needs to be managed; and, from the high level standpoint, access to the resources needs to be scheduled according to task deadlines. This paper describes a novel approach, based on hardware/software co-design modeling, to task allocation for scheduling tasks on a shared hardware under run-time constraints. The proposed task allocation technique has been implemented on a reconfigurable computing platform from Xilinxreg that employs Virtex-II protrade FPGA and embedded power PC core