Multi-level task partitioning algorithm based on heterogeneous signal processing platform

摘要

Aiming at the existing real-time task scheduling algorithm without considering the difference of node computing power in system heterogeneity, a multi-level task partitioning method for heterogeneous signal processing platform is proposed according to the characteristics of the real-time task. To achieve the task scheduling load balancing and low communication synchronization cost, the multi-level task partitioning algorithm based on data flow graphs draws on the idea of multi-level partitioning. We choose advanced telecom computing architecture as the experiment platform and the radar and other typical applications for test analysis. The experimental results show that the effectiveness of the algorithm.