The Research Of Hierarchical Clustering Algorithm With Gateway For Wireless Sensor Network

Wireless Sensor Networks (WSN) is currently closely watched at home and abroad, and has been the emerging multi-disciplinary high degree of cross-frontier areas of research, which combines the microelectronic technology, embedded computing technology, modern networking and wireless communication technology, distributed information processing technology and other advanced technology. And it can collaborative real-time monitor, sense and collect a variety of environmental or monitoring objects information in the network coverage area and then process them. Finally, the processed information is sent by the wireless means, and then it maybe be transported to the observer or the monitoring center via the self-organized multi-hop means of communication.In view of the inherent characteristics of wireless sensor networks, its routing protocol requirements are different from existing routing protocols in current communication networks. In particular, these nodes which are sowed in the circumstances in the one-time are non-renewable energy. So, how to extend the network lifetime of wireless sensor networks will naturally become the primary design goal, while the network topology is considered as an important platform of the routing protocols and MAC layer protocols. And it has been an objective basis for support how to make good use of this control. In this paper, we analysis the conclusion of the existing researches of topology control in wireless sensor network and then present a new algorithm of hierarchical structure of network topology in wireless sensor networks.This paper introduces the EAMCT-G (Energy-Aware Multilevel Clustering Tree with Gateway) algorithms, which adopts the concept of maximal independent set and minimum dominating set in graph theory. Then the paper proposes the local maintenance and fast updating of cluster-head. But there is a problem during the practical application and the current simulation:When the energy of a cluster-head is greater than the energy of other cluster-head in a proper range, it will put all the neighboring nodes into this cluster. So the number of the cluster'members will be so large in this cluster. And the cluster-head could be premature death due to energy consumption. So, the networks'lifespan will be much shorter. This paper improves on the EAMCT-G algorithm. If any nodes become cluster-head, some strategy will regulate these nodes and decide which nodes maybe join the member of the cluster. The strategy will make the entire network topology be much uniform in the number of cluster'members, much efficient on the transmission of data and much applicability in the actual large scale distribution of the nodes.