Situation Based Energy Management Framework on Mobile Devices

Chen Yeou Yu, Richard O. Oyeleke, Carl K. Chang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

With the prevalent usage of mobile phones, energy saving has become a critical issue for manufacturers. Extended battery life with increasing battery capacity is a common solution for hardware-based improvements. However, energy efficient management mechanism has attracted matching attention. In this paper, we present a situation based energy management framework. First, we collect data of activities of daily living (ADL) and apply deep learning for training. Next, Long Short Term Memory (LSTM) of Recurrent Neural Network (RNN) is used as our forecasting model to predict what a user will do at a certain hour in a certain day. Such prediction model relies on sensor fusion in gesture detection, camera-based facial detection and touch events on LCD screen in dynamic contextual environment. Finally, an inference engine is used to recommend actions to be takes for energy saving by digesting the information gathered on previous two stages, static information collecting and dynamic detection. In order to show the performance of our framework, we make use of external hardware as assistance tool which is suggested by Google developer's website.

Original languageEnglish
Title of host publicationProceedings - 2018 IEEE 42nd Annual Computer Software and Applications Conference, COMPSAC 2018
EditorsClaudio Demartini, Sorel Reisman, Ling Liu, Edmundo Tovar, Hiroki Takakura, Ji-Jiang Yang, Chung-Horng Lung, Sheikh Iqbal Ahamed, Kamrul Hasan, Thomas Conte, Motonori Nakamura, Zhiyong Zhang, Toyokazu Akiyama, William Claycomb, Stelvio Cimato
Pages284-289
Number of pages6
ISBN (Electronic)9781538626665
DOIs
StatePublished - 8 Jun 2018
Event42nd IEEE Computer Software and Applications Conference, COMPSAC 2018 - Tokyo, Japan
Duration: 23 Jul 201827 Jul 2018

Publication series

NameProceedings - International Computer Software and Applications Conference
Volume2
ISSN (Print)0730-3157

Conference

Conference42nd IEEE Computer Software and Applications Conference, COMPSAC 2018
Country/TerritoryJapan
CityTokyo
Period23/07/1827/07/18

Keywords

  • Activities of daily living (ADL)
  • Energy management
  • Facial detection
  • Gesture recognition
  • LSTM
  • RNN
  • Sensor fusion

Fingerprint

Dive into the research topics of 'Situation Based Energy Management Framework on Mobile Devices'. Together they form a unique fingerprint.

Cite this