A Context-Adaptive and Energy-Efficient Wireless Sensor Network for Debris Flow Monitoring

Abstract

Without the applicable methods of forecasting the outbreak of debris flows in a wide area as well as long time, wireless sensor networks (WSN) of monitoring debris flow are essential for protecting safety and property of human. In nowadays, a common and key consideration of the network would be the energy efficiency, since both of the soil moisture (SM) sensor and the radio are power-hungry, meanwhile, the node requires to react in several months by a battery power. In this article, we proposed an intelligent disaster forecasting system, which guarantees nodes working and sleeping periodically. With the system, the nodes are implemented by very low power consumption in sleep mode, the whole network reacts changes of environment with variable operating frequency. Since there is no proper wake-up sensor with ultra-low-power in the scene of debris flows, we adjust the sampling rate of node in terms of changes of the environment, which could be predicted by the Wavelet neural network. With these solutions, we extend the node's lifetime (more than 2 years), while keeping a reliable service simultaneously.