Power-efficient real-time scheduling based on multi-granularity resource reservation for multimedia services
- Authors
- Sun, Joohyung; Cho, Hyeonjoong
- Issue Date
- 8월-2017
- Publisher
- INST ENGINEERING TECHNOLOGY-IET
- Keywords
- real-time systems; scheduling; power aware computing; performance evaluation; Linux; resource allocation; multimedia computing; power-efficient real-time scheduling; multigranularity resource reservation; multimedia services; quality-of-multimedia services; QoMS; energy-restricted mobile devices; aggressive conservative low-power work demand analysis-with-multigranularity algorithm; lpWDA-MG-AGG; CON algorithm; power consumption minimisation; mobile devices; dynamic voltage scaling technique; rate monotonic scheduling policy; performance evaluation; Linux operating system; power metre
- Citation
- IET SOFTWARE, v.11, no.4, pp.171 - 180
- Indexed
- SCIE
SCOPUS
- Journal Title
- IET SOFTWARE
- Volume
- 11
- Number
- 4
- Start Page
- 171
- End Page
- 180
- URI
- https://scholar.korea.ac.kr/handle/2021.sw.korea/82655
- DOI
- 10.1049/iet-sen.2015.0108
- ISSN
- 1751-8806
- Abstract
- Recent advances in mobile technologies have led to improved quality of multimedia services (QoMS) in a variety of mobile devices. Because multimedia has become a major form of content consumption for mobile users, satisfying user expectation on QoMS in energy-restricted mobile devices is critical. This need has motivated us to develop an aggressive and conservative low-power work demand analysis with multi-granularity (lpWDA-MG-AGG/CON) algorithm, designed to minimise power consumption in mobile devices by utilising a dynamic voltage scaling technique while simultaneously ensuring QoMS based on a resource reservation scheme. In addition, the authors analytically showed the schedulability of the proposed scheme under the rate monotonic scheduling policy. For performance evaluation, the authors implemented the two lpWDA-MG algorithms and several existing algorithms in a Linux operating system. Specifically, the authors measured power consumption with a power metre and determined that the proposed algorithms consume about 40% less dynamic power than the other existing algorithms. Moreover, the authors found that the proposed algorithms ensure acceptable QoMS.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - Graduate School > Department of Computer and Information Science > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.