Research On Self-Organizing Technologies Of Wireless Sensor Networks

Integrated with sensor, embedded computer, distributed information processing,modern network and wireless communication technology, Wireless Sensor Networks(WSNs) merge information world and the physical world to change the interactionsbetween human being and nature. WSNs are important component technology of theInternet of things. In WSNs, the wireless sensor nodes construct network system in theform of self-organization to process information of monitored objects. This paper studiesthe self-organization technology of WSNs，and solves the networking of random nodes,formation of network boundary and quantity of network configuration commendably.The main content can be outlined as follows:Firstly, the brief history of WSNs is introduced, and then the application fields ofWSNs are summarized. This paper outlines the present situation of self-organizationtechnology of WSNs.Secondly, this paper induces specialities which the clustering and routing algorithmsof wireless sensor networks should have, basing on WSNs with other traditional wirelessnetworks, and then studies recent representative algorithms for the wireless sensornetworks. After that, it sums up the characteristics and the appropriate application, pointingout the limitation of them emphatically. The future trends of the algorithms are put forward.Thirdly, the performance of the overlapping clustering in uniform distribution of nodesis analysed, and then we overcome the shortage of the previous clustering algorithms to putforward the Self-organization Overlapping Algorithm for Clustering in Non-uniformdistribution (Nu-SOAC). By improving electing mechanism of cluster head and reducingenergy consumption of communication in the cluster, this paper proposes the improvedSelf-organization Overlapping Algorithm for Clustering (SOAC), which can deal withmany kinds of distribution. The results of simulations indicate that the algorithms achieveideal performance in uniform distribution and random distribution.Fourthly, essentially different from the previous clustering algorithms, whenconsidering residual energy of nodes in WSNs with random distribution, we propose aSelf-Organization Algorithm for Clustering Basing on Node Position and Connection Density (SACN) to generate clusters. On the basis of SACN, the Balancing ClusteringAlgorithm with Distributed Self-Organization (DSBCA) is come up with, with optimizingelecting mechanism of cluster head and decreasing energy consumption of communicationin the cluster. The performance of the novel algorithm is illustrated with a series ofsimulated tests, which indicate that the new algorithm can establish more balanceableclustering structure effectively and enhance the network life cycle obviously.Fifthly，the paper discusses the present situation of boundary search algorithm ofWSNs and points out the deficiency of previous studies. Different from former boundarysearch algorithm basing on single node, we carry out fusing clusters borderline to formboundary of WSNs based on overlapping clusters, then Moreover, the results ofsimulations points out that the algorithm generates the network boundary feasibly andreliably.Sixthly, basing on the wireless sensor network related research of the configuratingWSNs, in view of the node uniform distribution, we establishes the analysis model ofsingle hop and multiple hops communication respectively to calculate the basic demand ofnode energy and the optimal node communication radius. In the situation of the noderandom distribution, with the conclusion of quantitative analysis in single cluster expandedto the general case, the minimum of the network energy and optimal clustering radius arefigured out.Finally, the entire research is summarized, following with the future work discussions.