Towards a statically scheduled parallel execution of an FRP language for embedded systems

This paper presents an automatic parallelization method for a pure FRP language designed for embedded systems. Recent advances in microprocessor technology make it possible to use multicore processors even in resource-constrained embedded systems. To fully utilize such processors with a pure FRP language, we propose a method for parallelizing signal graph updates. The method is based on a static scheduling algorithm because it should run with a simple RTOS or without the help of operating systems. To evaluate the proposed method, we implemented a parallelizing compiler for XFRP, a pure FRP language for embedded systems. In this paper, we describe the parallelizing method and show some preliminary evaluation results using Linux-based systems that demonstrate the effectiveness of the method.