Research Of Data Aggregation In Wireless Sensor Networks

With the rapid development of sensor technology, computer network and wireless communication technology, wireless sensor networks(WSN) have been widely used in related fields. However, the quantity and species of the sensory data is increasing. Currently, that’s a heated researching topic that how to gather, transport and fuse the sensory data effectively. Additionally, as sensor nodes in complex network environment are limited in power, bandwidth, memory and computational capacities, for which it’s necessary to fuse the sensory data and save the energy consumption and prolong the lifetime of networks.This paper emphatically describes the advantages and disadvantages of the existing data aggregation theory based on introducing wireless sensor networks firstly and does the related research as follows in the foundation of analyzing and studying the virtual grids model and loop structure..1. Research on the clustering algorithms of WSN and the scheduling algorithms of the cluster heads in each cluster in the process of data collection based on the loop structure in the virtual grids model.2. Analyzes the reason of the data redundancy and the essence of the spatio-temporal correlation and the method of representation of data, and then put forward the method for dealing with the correlation.3. Proposes a model of distributed data aggregation in wireless sensor networks based on wavelet transform, which improves the loop structure and the scheduling algorithm of cluster head in the process of data collection and then put the data validation mechanism into the data collecting process. Finally, fuse the sensory data by the wavelet transform theory.4. Raises a data aggregation algorithm based on data correlation clustering, firstly, define and compute the correlation of data collected by sensor nodes in different groups and then analyze the overall distribution by clustering algorithm. Compute the best aggregation point by joint probability density function to check whether the aggregation results are good or not. The algorithm also can be used to check whether the monitoring target is in a good state or not.In order to prove the efficiency and robustness of the model and the algorithm, this paper does experiment by network simulator version 2 and the simulation results conclude the model and algorithm is better than the related ones.