Name
Abstract | ||
|---|---|---|
Embedded systems execute applications that exercise the hardware differently depending on the computation task, generating varying workloads with time. Energy minimization can be reached exploring the optimal CPU frequency for each workload. We propose an autonomous and online approach, capable of minimizing energy through adaptation to these workload variations even in an unknown environment. In the proposed approach we use a reinforcement learning algorithm that suitably selects the appropriate CPU frequency based on workload predictions to minimize energy consumption. The proposed approach is validated through simulation using real smartphone data, an ARM Cortex A7 processor used in a commercial smartphone with Android 4.4.4 version was employed. Our proposed approach demonstrated to have an improvement in the Q-learning cost function and can effectively minimize energy consumption by up to 29% compared to the existing approaches. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.1109/DSD.2017.68 | 2017 Euromicro Conference on Digital System Design (DSD) |
Keywords | Field | DocType |
energy reduction,energy saving,mobile energy saving,machine learning,reinforcement learning,non-linear power predictor | ARM architecture,Power management,Central processing unit,Android (operating system),Computer science,Workload,Real-time computing,Minification,Energy consumption,Energy minimization,Embedded system | Conference |
ISBN | Citations | PageRank |
978-1-5386-2147-9 | 1 | 0.35 |
References | Authors | |
5 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Sidartha A. L. Carvalho | 1 | 1 | 1.70 |
| D. C. Cunha | 2 | 5 | 4.46 |
| Abel Guilhermino Silva-Filho | 3 | 62 | 12.94 |