Design of Fast and Efficient Energy-Aware Gradient-Based Scheduling Algorithms for Heterogeneous Embedded

In this paper, we present two heuristic energy-aware scheduling algorithms—1) Energy Gradient-based Multiprocessor Scheduling (EGMS) algorithm and 2) Energy Gradient-based Multiprocessor Scheduling with Intratask Voltage scaling (EGMSIV) algorithm—for scheduling task precedence graphs in an embedded multiprocessor system having processing elements with dynamic voltage scaling capabilities. Unlike most energy-aware scheduling algorithms that consider task ordering and voltage scaling separately from task mapping, our algorithms consider them in an integrated way. EGMS uses the concept of energy gradient to select tasks to be mapped onto new processors and voltage levels. EGMSIV extends EGMS by introducing intratask voltage scaling using a Linear Programming (LP) formulation to further reduce the energy consumption. Through rigorous simulations, we compare the performance of our proposed algorithms with a few approaches presented in the literature. The results demonstrate that our algorithms are capable of obtaining energy-efficient schedules using less optimization time. On the average, our algorithms produce schedules which consume 10 percent less energy with more than 47 percent reduction in optimization time when compared to a few approaches presented in the literature. In particular, our algorithms perform better in generating energy-efficient schedules for larger task graphs. Our results show a reduction of up to 57 percent in energy consumption for larger task graphs compared to other approaches. Index Terms—Energy-aware scheduling, power management, dynamic voltage scaling, energy minimization, heterogeneous multiprocessor scheduling, embedded systems. C