A Kind of Space-Time Event Detection Approach for Wireless Sensor Network

Space-time event detection is one of major tasks in wireless sensor network. The event always represents the complex space-time relationship. Despite there are great progress in this area, most of the existing works on event detection are restricted to inference and study in the spatial or the temporal field separately, the accuracy of event detection is not so high. Based on the description a space-time event, a space-time model based on dynamic conditional random field is given and an algorithm for space-time event detecting is presented based on belief propagation in this paper. Simulation results show that the presented algorithm is better than the algorithm in performance based on Markov random field and dynamic Bayesian network on detection accuracy, recall rate, and F1 score, the proposed algorithm is suited for wireless sensor network with short timeconsuming. Copyright © 2013 IFSA.