An Energy-Saving Algorithm in Networked Embedded System Based on Critical Tasks Served First Strategy

Abstract

Lots of collaborative applications that contain a number of dependent tasks run in networked embedded system, such as wireless sensor network, Cyber-Physical-System and so on. For the reasons that deadlines of some collaborative applications are usually not so hard and lifetime of whole system depends on energy dissipation, it is meaningful to figure out a good strategy for balancing delay with system energy dissipation. Considering this issue, our research focuses on trying to minimize total system energy dissipation for extending lifetime of whole system, on the premise that delay of finish time is acceptable. In this paper, we point out which tasks are critical among independent tasks from the perspective of energy-saving and propose BESF (Biggest Energy Span First) algorithm that first allocates the task with biggest energy span. Actually, BESF algorithm references BEATA algorithm, however, BESF could not only further decrease energy-consumption of whole system comparing with existing related algorithms, like BEATA, GEATA, but also still ensure that the finish time is tolerable. Experimental results show that in most situations, BESF algorithm is able to save more energy comparing with existing algorithms.