Compressive Sensing for Remote Flood Monitoring

Although wireless sensor networks (WSNs) are considered as one of the prominent solutions for flood monitoring; however, the energy constraint nature of the sensors is still a technical challenge. In this paper, we tackle this problem by proposing a novel energy-efficient remote flood monitoring system, enabled by compressive sensing. The proposed approach compressively captures water level data using; i) a random block-based sampler, and ii) a gradient-based compressive sensing approach, at a very low rate, exploiting water level data variability over time. Through extensive experiments on real water-level dataset, we show that the number of packet transmissions as well as the size of packets are significantly reduced. The results also demonstrate significant energy reduction in sensing and transmission. Moreover, data reconstruction from compressed samples are of high quality with negligible degradation, compared to classic compression techniques, even at high compression rates.