Energy management for multiple real-time workflows on cyber-physical cloud systems

摘要

Cyber-physical cloud systems (CPCS) are extensions of cyber-physical systems (CPS) that expand the cyber-part and distribute it on-device and in-cloud. CPCS are considered large-scale heterogeneous distributed cloud computing systems that support execution of multiple workflows. This study aims to reduce the energy consumption of multiple real-time workflows on CPCS and it contains two objectives: (1) maximizing the number of workflows that are completed within their deadlines; (2) minimizing the energy consumption of the workflows that ate completed within their deadlines. The former is solved by proposing a deadline-driven processor merging for multiple workflows (DPMMW) algorithm, whereas the latter is solved by proposing a global energy saving for multiple workflows (GESMW) algorithm to minimize the total energy consumption. Experimental results validate that the combined DPMMW&GESMW algorithm can reduce deadline miss ratio (DMR) and save as much as possible energy over existing methods.