Sensor selection using circular target model in Visual Sensor Networks

This paper is a study on sensor selection problem in Visual Sensor Networks (VSNs). The objective of this problem is finding an optimal subset of visual sensor nodes witch cover all of the targets such that overall network cost is minimized. In addition, in order to efficiently utilize of the bandwidth, the correlation of data which are calculated based on captured data of sensors should be minimized. For this achievement, a circular model for target modeling is proposed and furthermore, a criterion is defined to measure the correlation of the captured data from sensors. Experimental results demonstrate that applying our model for Full / Partial Angle Coverage problem compared to K-Coverage problem redound the less correlated sensor nodes.