Research On Clustering Algorithm Based On Cooperative Nodes

In wireless sensor networks(WSNs), hundreds of low-cost sensors with limited battery energy are usually scattered in the monitoring area. These nodes are self-organized into a wireless network and periodically report their data to the sink. Due to the energy limitation and non-recharging of nodes, sensor nodes are easily out of work because their energy is exhausted. Therefore, how to gather data efficiently while prolonging the network lifetime is one of challenges in resource-constrained WSNs.The clustering routing algorithms such as LEACH(Low-energy Adaptive Clustering Hierarchy) can effectively improve the lifetime of WSNs. Clustering usually has two processes: selecting the nodes with more residual energy as cluster heads and rotating cluster heads periodically. Cluster heads tend to consume much more energy due to data gathering and relaying. For solving these problems, we propose a centralized clustering algorithm and a distributed clustering algorithm based on cooperative nodes after analyzing existing clustering routing algorithms. The contributions of this thesis are as follows:(1) In this thesis, we propose an analytical model to investigate the optimal number of clusters in a wireless sensor network. As the number of clusters increases, the energy consumption of intercluster transmissions will increase remarkably. But with the decrease of the number of clusters, the energy consumption of intracluster transmissions will increase. Therefore, it is of significance to obtain the optimal number of clusters. In this energy model, by using cooperative nodes, we can not only reduce the energy consumption of cluster heads efficiently but also prolong the lifetime of network.(2) We propose a centralized clustering method based on spectral partitioning. Due to the limitations of centralized clustering algorithm, we propose a distributed clustering algorithm based on fuzzy c means.(3) To balance the energy consumption among sensor nodes, we propose an efficient strategy to choose cooperative nodes and cluster heads. During selecting cluster heads and cooperative nodes, we not only take into account of remaining energy but also some important factors such as location, which make energy consumption more balanced.Finally, we conduct extensive simulations, and the results show that the proposed algorithms outperform the hybrid energy-efficient distributed(HEED) clustering algorithm in terms of energy cost and network lifetime.