The Data Aggregation Algorithm Of Wireless Sensor Networks

The nodes in wireless sensor networks are random distributed in the monitoring area, which is characterized by dense deployment of nodes. The data from the close nodes have a lot of similarity. When the nodes transmit the data directly to the Sink node, it will result in a large amount of data redundancy and network energy consumption. So that the similarity of the data is not need to be all sent to the Sink node, we have to aggregate these data. The pretreatment of the data should be performed before data aggregation, remove some similarity data. On the premise of guarantee the data accuracy, it can reduce the amount of data transmission in the wireless sensor network, and then extend the life of the network.In this paper, we propose different solutions based on the study of data preprocessing and data aggregation method, as shown below:(1)The paper proposes an uneven clustering aggregation method based on the data density correlation degree, aiming at solving the “hotspot” problem(Energy consumption of the network is fast and the life of it is short) in the WSNs. The method consists of two stages: in the first stage, the paper selects representative nodes and isolated nodes according to the density correlation degree of nodes, and in the second stage, the high level data aggregation is performed on the selected nodes. That is, cluster heads are selected using the uneven clustering algorithm and proceeds with the data aggregation which are sent to the Sink node through multi hop. The simulation results show that this method can prolong the lifetime of the network effectively using the network energy equilibrium.(2)Also based on the problem of "hot spots" in the network we propose a kind of data pretreatment method based on statistical model, namely distribution display method, and then according to the cluster cooperation link algorithm for data aggregation. This method is divided into two stages: in the first stage, the algorithm calculate the intermediate of the data and divide the data into upper part and the lower part, then calculate the median, and obtain the effective criterion of interval. In the second stage, according to the residual energy of node and the distance between the neighbor nodes and the number of neighbor nodes to select the cooperative node and cluster head. The cluster head node only collects and fuse data, and the cooperative node sends the aggregated data to the base station through the link. This method can effectively balance the network energy and prolong the network lifetime.