A Communication-aware Scheduling Algorithm for Hardware Task Scheduling Model on FPGA-based Reconfigurable Systems

Task scheduling is an important aspect of high performance reconfigurable computing. Most of the heuristics for this NP-hard problem are based on a simple abstract model of FPGA and have little investigation into optimizing data communication which influences the system performance importantly. To solve this problem, a Communication-aware Maximum Adjacent Edges (CA-MAE) algorithm based on new 2D reconfigurable model is proposed, which could reduce communication distance during scheduling and enhance the system performance. The experimental results show that CA-MAE reduces communication cost by 17%.