Distributed detection in landslide prediction based on Wireless Sensor Networks

Wireless sensor network is a newly technology t be used in landslide prediction as they provide dense data, and real-time monitoring and prediction of events. A real-time monitoring system of landslide prediction based on WSN is introduced in this paper. Statistical modeling of the landslide strain data and the distributed detection algorithm are mainly analysed. The strain data are modeled using variable mean of Gaussian process. Miss alarm ratio is a critical performance parameter to landslide prediction. Comparing with centralized detection method, simulation result shows that distributed detection algorithm performs better than centralized detection in criteria of miss alarm ratio and wrong alarm ratio of landslide.