Multi-sensor Data Fusion Using Weighting Method based on Event Frequency
Donghyok Suh, Chang-Keun Ryu · The Journal of the Korea institute of electronic communication sciences · 2011
A wireless sensor network needs to consist of multi-sensors in order to infer a high level of information on circumstances. Data fusion, in turn, is required to utilize the data collected from multi-sensors for the inference of information on circumstances. The current paper, based on Dempster-Shafter's evidence theory, proposes data fusion in a wireless sensor network with different weights assigned to different sensors. The frequency of events per sensor is the crucial element in calculating different weights of the data of circumstances that each sensor collects. Data fusion utilizing these different weights turns out to show remarkable difference in reliability, which makes it much easier to infer informaiton on circumstances.