Research Of Clustering Algorithm In Wireless Sensor Network

With the rapid development of technology advances and continues mature of skill, the same with the rising demand of human discovery for natural transformation of nature, wireless sensor networks for its specific application scenarios as an effective tool for quick access to the natural human discovery of human come into people’s vision. Meanwhile, continues mature of LSI integration technology and rapid miniaturized processors also come out to establish a solid foundation for the hardware of large-scale wireless sensor networks. The application and promotion of wireless sensor networks will occupy increasingly important position in human life.However, the study found that current wireless sensor networks has a very uneven energy consumption phenomenon:on the one hand, it dues to the small size of sensor nodes, the energy carried by extremely limited, on the other hand, the various nodes in the network is different because the division of labor, energy consumption exists a big difference. The two factors greatly affect the network lifetime, resulting in a partial energy network failure, or energy appears empty, shortening the life of the network. Research in this field around energy and network lifetime two timeless themes, so how to achieve efficient use of wireless sensor network nodes energy and research how to extend the network lifetime problem has been the academic and industry widespread attention, while research in these two areas also be the problem need to be overcome. The existing traditional flat topology is unable to implement robust and efficient network resource management, data collection, etc. Wireless sensor network clustering mechanism is similar to the human management structure, can be fully used to manage the energy and resources in the wireless sensor network (WSN), and can achieve maximum valid data collection monitored by the senor nodes.This article focuses on how to select the appropriate cluster head and the establishment of cluster structure. The research shows that most of the existing clustering scheme does not consider the residual energy, though they have low complexity and easy to implement, but caused the uneven network load and unequal energy consumption. Therefore, this paper first studies the status of wireless sensor networks, from the existing classical clustering algorithm puts forward a new cluster head election algorithm based on joint probability of comprehensive weights(JPCW). In the JPCW, introducing the theory of joint probability and standard deviation method in statistical are introduced, and construct a based on joint probability of comprehensive weight model. On the basis of this model, each node itself situation and the whole network performance analysis are given. The algorithm steps, theoretical principle, and the algorithm performance are described in detailed. At last, proves the validity of the algorithm through simulation. This algorithm prolongs the network life and achieves a balance network load.