Application Of Rough Fuzzy Sets Of Data Fusion In Sensor Network

With the extensive application of computer technology and wireless sensornetworks in people’s lives, it is hoped that the dialogue with the surroundingenvironment is more natural and direct. Meanwhile, the implementation monitoringsensor networks will inevitably need to transfer and process large amounts ofinformation from which there are a lot of redundancy, uncertain information. Datafusion technology appears in order to delete unnecessary errors and redundantinformation from a large number of information. Though data fusion technology hasnot yet formed a perfect system structure, but the application in specific areas appearsmuch and achieved better results.There are many similarities and compatibilities between rough set and fuzzy set twhile dealing with uncertain problems, but they concern different parts of a problem,and they also have strong complementary characteristics, then rough fuzzy set theoryappears. The main purpose of the paper is to design a data fusion algorithm whichbases on rough fuzzy set theory, then to verify its validity through a large number ofdata which is collected by sensor network. The other two data fusion algorithms thatbase on fuzzy set and rough set are also designed in the paper as the comparison.Firstly, the paper analyzed the ideology of three algorithms in detail and pointedout the difficulties, advantages and disadvantages when used in data fusion. Secondly,we designed the algorithms steps and implementation process on the basis of the threekinds of thoughts. Finally, we use the abundant data coming from sensor network tomake fusion reasoning with three fusion methods.It possible to draw the following conclusion from the comparison and analysisof results using three kinds of fusion algorithm, Compared with fuzzy set dataalgorithm, the membership function which based on rough fuzzy set is doesn’t totallydepended on expert knowledge any longer, and all reasoning are based on the data,which can improve the objectivity, reliability of the fusion results. The algorithmenhances the predictive ability of observation object while comparing to rough setmethod, and there was a marked improvement in the processing of multiple attributedata. In short, in the process of this study, the virtue of rough fuzzy sets has beenembodied, and the data fusion instances show that it better than a single theory.