DISTRIBUTED DETECTION STRATEGIES FOR LANDSLIDE PREDICTION USING WIRELESS SENSOR NETWORKS

This paper, proposes various distributed detection algorithms for landslide prediction using Wireless Sensor Network. Sim- ulation results show that statistical detection methods outper- form traditional threshold based detection in low SNR condi- tions. This allows us to use inexpensive sensors with low sen- sitivity. Among the proposed statistical detection strategies, Distributed Vector Based Detection with Independent Nodes (DVBD-IN) outperforms others in terms of having the lowest false alarm rate. Nevertheless, it has high energy consump- tion.