Timing Prediction for Dynamic Application Migration on Multi-core Embedded Systems

Zheng Li, Hao Wu, Shuibing He

Abstract: To accommodate execution mode change and hardware malfunction, dynamic system reconfiguration, which invokes application migration across different processing cores, needs to be supported on multi-core embedded systems. Different application migration strategies will impact system's timing behaviors in different manners, it is important to select an appropriate one such that the system's timing performance after the migration process is still acceptable. The focus of our research is to predict the system's timing change of possible migration strategies and upon which to choose the optimal one. Extensive experiments have been set up by running multiple benchmarks and experimental results validate the effectiveness of our proposed approach.