Research Of Clustering Algorithm Based On Data Content For Heterogeneous Sensor Networks

In recent years, with the progress of MEMS technology, sensor technology and wireless communication technology, the development of low-power multi-sensor is promoted sharply so that it will be able to integrate within a small volume of information collection, data processing and wireless communications and other functions. Wireless sensor network is composed of a large number of minisensor nodes which is deployed in monitoring region to form a multi-hop's self-organizing network system through wireless communication, which can percept, acquire and process monitoring information of network coverage area, and send it to observers. It is difficult to add energy because of the limitation on environmental constraints while it is generally rely on battery-powered sensor nodes, therefore, how to effectively conserve energy to maximize the network life cycle is the first wireless sensor network design problem.In this paper, we first introduce the concepts, structures, features, key technologies and application fields of WSN briefly, and then analyze and compare the routing protocols for WSN which are mainly used.As for common existing and application need of heterogeneous wireless sensor networks in current practical applications, we propose a clustering algorithm CADC for heterogeneous sensor networks that base on data content after consideration upon initial energy difference of heterogeneous wireless sensor networks and information collection. In the CADC algorithm, Networks cluster according to data contents collected by the nodes, which means clustering the same type of nodes in order to achieve the optimal data fusion rate, and the selection of cluster head is based on comprehensive survey of the residual energy, distance, and energy consumption rate and other factors to achieve load balancing purposes. Finally we do the comparison between the algorithm put forward in the paper and the LEACH algorithm through simulation experiment using the simulation platform OMNET++.The results show that our CADC algorithm owns better performance and may extend the network lifetime effectively. It applies scenarios that collecting multiple data.