摘要：Since the measuring accuracy and environment of each sensor are different, there must exist difference in the correctness of measurement. If the testing data is not processed and utilized with distinction, it will cause impreciseness to the testing results and lead to errors of the system. It is necessary to selectively distinguish the importance among the sensors, contraposing the situation of each sensor in the testing system and the accuracy of tests. So the related concepts of data aggregation technology in wireless sensor networks and the aggregation algorithm performance evaluation criteria are introduced. The core problem in WSN, aggregation operation for sensing data, is studied deeply. The problems in node data group when the distributed clustering technology is implemented to WSN are also analyzed. Then a distributed K-mean clustering algorithm based on WSN is proposed. On the basis of this improved algorithm, we realize a network data aggregation processing mechanism based on adaptive weighted allocation of WSN. DKC algorithm is mainly used to process the testing data of bottom nodes. When reducing the data redundancy it can provide more accurate field testing information and system status information. It can make rapid packet for the network nodes. The packed data will be used to provide correct judgement, according to the size of its corresponding weight, to acquire more reasonable results. The experiments have demonstrated that our method can greatly decrease the data redundancy of WSN and save large amount of storage resources. The network bandwidth consumption is also reduced. So this scheme has high efficiency and good scalability.
