Task offloading technique using DMIPS in wearable devices

Wearable devices such as smart watches become popular, especially thanks to connectivity with smartphones. They are, however, sensitive to energy consumption that is caused by computation and networking capabilities. This motivates to develop task offloading techniques that enable to perform task computation on smartphones rather than on wearable devices. In this paper, we propose a new offloading method by adopting a common factor that determines the amount of offloading work. Our method prevents unintentional offloading through the architecture using DMIPS that determines offloading workload which can be commonly used in different hardware systems. We also compared the energy consumption of smart watches with and without offloading at the application level.