Journal
IEEE SENSORS JOURNAL
Volume 16, Issue 17, Pages 6793-6804Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSEN.2016.2587220
Keywords
Wireless sensor networks; energy harvesting; prediction algorithms; solar energy; wind energy
Funding
- Sapienza Universita di Roma through the Harvesting-Aware Communication Protocols in the Self-Powered Internet of Things Project
- Ministero dell'Istruzione, dell'Universita e della Ricerca through the Smartour (Intelligent Platform for Tourism) Project
- Ministero dell'Istruzione, dell'Universita e della Ricerca through the Social Museum and Smart Tourism Project [CTN01000342315]
Ask authors/readers for more resources
The increasing popularity of micro-scale power-scavenging techniques for wireless sensor networks (WSNs) is paving the way to energy-autonomous sensing systems. To sustain perpetual operations, however, environmentally powered devices must adapt their workload to the stochastic nature of ambient sources. Energy prediction models, which estimate the future expected energy intake, are effective tools to support the development of proactive power management strategies. In this paper, we present profile energy prediction model (Pro-Energy), an energy prediction model for multi-source energy-harvesting WSNs that leverages past energy observations to forecast future energy availability. We then propose Pro-Energy with variable-length timeslots (Pro-Energy-VLT), an extension of Pro-Energy that combines our energy predictor with timeslots of variable lengths to adapt to the dynamics of the power source. To assess the performance of our proposed solutions, we use real-life solar and wind traces, as well as publicly available traces of solar irradiance and wind speed. A comparative performance evaluation shows that Pro-Energy significantly outperforms the state-of-the-art energy predictors, by improving the prediction accuracy of up to 67%. Moreover, by adapting the granularity of the prediction timeslots to the dynamics of the energy source, Pro-Energy-VLT further improves the prediction accuracy, while reducing the memory footprint and the energy overhead of energy forecasting.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available