Scheduling algorithms for energy minimization
Sanjay Ranka, Sartaj K. Sahni, Jaeyeon Kang · 2008
Energy consumption is a critical issue in parallel and distributed embedded systems. We present novel algorithms for energy efficient scheduling of DAG (Directed Acyclic Graph) based applications on DVS (Dynamic Voltage Scaling) enabled systems. The proposed scheduling algorithms mainly consist of assignment and slack allocation. All schemes for the assignment and the slack allocation effectively minimize energy consumption while meeting the deadline constraints in static or dynamic environments. They are also equally applicable to the homogenous and heterogeneous parallel machines. Experimental results show that the proposed algorithms provide significantly good performance for energy minimization and require considerably small computational time.