Clustering Data Fusion Algorithm Based on D-S Theory for Wireless Sensor Networks
Shaoqiang Liu · Microcomputer Information · 2009
In wireless sensor networks, the recognition results of multi-sensor for the same goal are often inaccuracy due to external interference and internal error. As a non-precise reasoning method in the field of identification, the D-S theory is effective for the data fusion in multi-sensor to make a reasonable decision. In this paper, on the basis of previous studies, a clustering data fusion method based on D-S theory for WSN is proposed. In this method, the information entropy of evidence and the distance between evidences are used to select cluster heads and identify cluster members and candidate evidences. This method can decrease the redundant information transmission in clusters, reduce the computational and improve the accuracy of data fusion in cluster heads, thus provide more reliable recognition information about the type and location of targets for users.