Distributed Event Detection in Sensor Networks under Random Spatial Deployment

• Main Authors: P. Zhang, G. Peters, I. Nevat, G. Xiao, TAN, Hwee-Pink
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2014
• Subjects: Event detection; Finite Binomial Point Process; Gram-Charlier series expansion; Sensor networks; Digital Communications and Networking
• Online Access: https://ink.library.smu.edu.sg/sis_research/2942; https://ink.library.smu.edu.sg/context/sis_research/article/3942/viewcontent/milcom2014.pdf
• Institution: Singapore Management University

We present a novel event detection algorithm in sensor networks for the case where the sensors are randomly deployed in space. In particular we consider a random sensors deployment according to a Homogeneous Finite Binomial Point Process. We first derive the optimal event detection decision rule. We then develop a novel algorithm to evaluate the intractable marginal likelihood based on the Gram-Charlier series expansion. We evaluate our algorithms through extensive Monte Carlo simulations. Simulation results present the detection and false alarm rates for different system parameters such as number of sensors deployed, deployment region size etc.