Title
Real-time loop scheduling with energy optimization via DVS and ABB for multi-core embedded system
Abstract
Dynamic Voltage Scaling (DVS) is an effective technique to reduce energy consumption of processors by dynamically adjusting the operational frequency and supply voltage. However, with feature sizes shrinking, the achievable power saving by DVS is becoming limited as the leakage power increases exponentially. Adaptive Body Biasing (ABB) is an effective technique to reduce leakage power by increasing the circuit's threshold voltage via body biasing. In this paper, we propose a novel real-time loop scheduling technique to minimize both dynamic and leakage energy consumption via DVS and ABB for applications with loops considering voltage transition overhead. The proposed algorithm, EOLSDA (Energy Optimization Loop Scheduling with DVS and ABB), is designed to repeatedly regroup a loop based on rotation scheduling [4,5] and decrease the energy consumption via DVS and ABB within a timing constraint. We conduct experiments on a set of DSP benchmarks based on the power model of 70nm technology. The results show that our technique achieves big energy saving compared with list scheduling [8] and the algorithm in [11].
Year
DOI
Venue
2007
10.1007/978-3-540-77092-3_1
EUC
Keywords
Field
DocType
power model,leakage power increases exponentially,energy consumption,multi-core embedded system,list scheduling,effective technique,energy optimization,leakage power,novel real-time loop scheduling,achievable power saving,big energy,leakage energy consumption,real time,threshold voltage,embedded system
Dynamic voltage scaling,Digital signal processing,Scheduling (computing),Computer science,Voltage,Real-time computing,Loop scheduling,Multi-core processor,Energy consumption,Embedded system,Energy minimization
Conference
Volume
ISSN
ISBN
4808
0302-9743
3-540-77091-7
Citations 
PageRank 
References 
3
0.41
10
Authors
5
Name
Order
Citations
PageRank
Guochen Hua130.41
Meng Wang2808.46
Zili Shao31618134.03
Hui Liu41269.88
Chun Jason Xue51616140.95