No abstract

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

This paper presents a power management technique for improving the efficiency of harvesting energy from air-flows in wireless sensor networks (WSNs) applications. The proposed architecture consists of a two-stage energy conversion circuit: an ac-dc converter followed by a dc-dc buck-boost regulator with maximum power point tracking capability. The key feature of the proposed solution is the adaptive hybrid voltage rectifier, which exploits both passive and active topologies combined with power prediction algorithms. The adaptive converter significantly outperforms other solutions, increasing the efficiency between 10% and 30% with respect to the only passive and the only active topologies. To assess the performance of this approach in a reallife scenario, air-flow data have been collected by deploying WSN nodes interfaced with a wind microturbine in an underground tunnel of the Metro B1 line in Rome. It is shown that, using the adaptive ac-dc converter combined with power prediction algorithms, nodes deployed in the tunnel can harvest up to 22% more energy with respect to previous methods. Finally, it is shown that using power management techniques optimized for the specific scenario, the overall system overhead, in terms of average number of sampling performed per day by a node, is reduced of up to 93%.

No abstract

Structural health monitoring is a vital tool to help engineers improving the safety of critical structures, avoiding the risks of catastrophic failures. Wireless sensor networks (WSNs) are a very promising technology for structural health monitoring, as they can provide a quality of monitoring similar to conventional (wired) SHM systems with lower cost. In addiction, WSNs are both non-intrusive and non-disruptive and can be employed from the very early stages of construction.The main goal of this work is to investigate the feasibility of a WSN with energy-harvesting capabilities for structural health monitoring, specifically targeting underground tunnels

The increasing popularity of micro-scale energy-scavenging techniques for wireless sensor networks (WSNs) is opening new opportunities for the development of energy-autonomous systems. To sustain perpetual operations, however, environmentally-powered motes must adapt their workload to the stochastic nature of ambient power sources. Energy prediction algorithms, which forecast the source availability and estimate the expected energy intake in the near future, are precious tools to support the development of proactive power management strategies. In this work, we propose Pro-Energy-VLT, an enhancement of the Pro-Energy prediction algorithm that improves the accuracy of energy predictions, while reducing its memory and energy overhead

No abstract