A novel spatial correlation model for wireless sensor network applications

Wireless sensor networks (WSN) are densely deployed to promise the fine-grain monitoring in various applications. For example, it may be as high as 20 nodes/m³ or more [1]. Due to high density of sensor nodes, spatially correlated information is observed and transmitted by surrounding sensor nodes once an interest of event detected. Thus, there exists spatial correlation among the sensor observations. The spatial correlation brings significant potential advantages along with collaborative nature of the WSN in energy-efficient design of communication protocols. This paper presents a novel spatial correlation model for wireless sensor networks. Based on sensor coverage model and location of sensor nodes, a spatial correlation function is derived to describe the correlation characteristics of measurements observed by sensor nodes. The case studies using correlation function are performed to study the correlation relationship between sensor nodes. Finally, based on case studies, their results, and discussions, a correlated cell construction algorithm is proposed and possible approaches are explored to exploit spatial correlation for efficient medium access and clustering protocols for WSN.