Treffer: An efficient battery-aware task scheduling methodology for portable RC platforms

Title:
An efficient battery-aware task scheduling methodology for portable RC platforms
Source:
FPL 2004 : field-programmable logic and applications (Antwerp, 30 August - 1 September 2004)Lecture notes in computer science. :669-678
Publisher Information:
Berlin: Springer, 2004.
Publication Year:
2004
Physical Description:
print, 15 ref
Original Material:
INIST-CNRS
Document Type:
Konferenz Conference Paper
File Description:
text
Language:
English
Author Affiliations:
Department of ECECS, University of Cincinnati, Cincinnati, OH 45221-0030, United States
ISSN:
0302-9743
Rights:
Copyright 2004 INIST-CNRS
CC BY 4.0
Sauf mention contraire ci-dessus, le contenu de cette notice bibliographique peut être utilisé dans le cadre d’une licence CC BY 4.0 Inist-CNRS / Unless otherwise stated above, the content of this bibliographic record may be used under a CC BY 4.0 licence by Inist-CNRS / A menos que se haya señalado antes, el contenido de este registro bibliográfico puede ser utilizado al amparo de una licencia CC BY 4.0 Inist-CNRS
Notes:
Computer science; theoretical automation; systems

Electronics
Accession Number:
edscal.16107539
Database:
PASCAL Archive

Weitere Informationen

In this paper we present a simple yet efficient methodology for battery-aware task execution on FPGAs in portable Reconfigurable Computing (RC) platforms. We divide the reconfigurable area on an FPGA into several fixed reconfigurable slots called Configurable Tiles. We then schedule real-time tasks onto these tiles. Various schedules using different number of tiles are calculated off-line. These schedules along with their execution times are then sent to a run-time scheduler which dynamically decides, which schedule is the most battery efficient. By varying the number of tiles used for scheduling tasks, we can vary the battery usage and lifetime. We tested the methodology by running it on several different task graph structures and sizes, and report an average of 14% and as high as 21%, less battery capacity used, as compared to non-optimal execution. Finally, we present a case study where we implement a real-time face recognition algorithm on the iPACE-V1 [6] platform using the proposed methodology and observed 1.3 to 3.3 times improvement in battery life-time.