A New Classification Method on Information Fusion of Wireless Sensor Networks
Chanchan Zhao, Yuxing Wang · 2008
Information fusion still faces many problems in WSN (wireless sensor networks). So far, uniform and accurate definition about information fusion of WSN is rare. Because WSN provides service to the users, we take userpsilas requirements as a springboard. To analyze common information fusion definitions and considering adequately the most important factor-userpsilas requirements, then, a novel definition about information fusion of WSN appeared; hereby an original method of information fusion classification in view of userpsilas requirements based on SOM (self-organizing feature map) is given. SOM is powerful in processing magnanimous clustering information. It is the point that we take it as one infusion method in our structure. Experiments were done proved that different methods fitted different userpsilas requirements to make user satisfied. The advantage is that such classification and disposal method according with particular characters of WSN not only can obtain more exact and more general information in application but also can assist in intelligence decision-making later.