Wireless Sensor Network Clustering Algorithm Research

A wireless sensor network (WSN) consists of thousands of low-cost, low-power and tiny sensor nodes. Sensors in WSNs gather information from the environment and then transmit the gathered data to the base station (BS). The BS processes the data and sends the processed data to the external network.Sensors are usually battery-powered. Limited by the low-cost and small-size, the energy resources of sensors are limited and non-rechargeable. Therefore, a major objective of protocol design for WSNs is maximizing the network lifetime. Clustering technology groups the nodes into clusters and limits the communication among cluster heads which can save a lot of energy of nodes and prolong the network lifetime.Based on the analysis of a large number of existing works, three protocols are proposed in this thesis. We give the theoretical analysis of our protocols and simulate them in NS2 or MATLAB. Theoretical analysis and simulation results of each protocol show that these protocols can obtain a good cluster head distribution and increase the network lifetime significantly.This thesis includes six chapters. The first chapter introduces the concept, system structure, network characteristics and application type of wireless sensor networks. The second chapter has a simple overview of the basic theory of clustering algorithms for wireless sensor networks, and categories some of the existing classical clustering algorithms. In the third chapter, an energy-efficient distributed clustering algorithm called EEDC is proposed. It iteratively chooses sensor nodes with high residual energy as tentative cluster heads to participate in the final cluster head competition. Furthermore, EEDC elects tentative cluster heads to be final cluster heads by a novel probability which is based on the intra-cluster communication cost. Finally, the cluster heads generated by EEDC are nodes with high residual energy and low intra-cluster communication cost, which can significantly increase the network lifetime. In the fourth chapter, based on the wireless sensor networks with non-uniform node distribution, a cluster-based routing protocol (CBRP) is proposed. CBRP is composed by an energy-aware clustering algorithm EADC and a cluster-based routing algorithm. EADC constructs clusters in a competition way. At the same time, the routing algorithm increases forwarding task of the nodes in scarcely covered areas by forcing cluster heads to choose nodes with more energy and less member nodes as its next hop. Finally, achieves load balance among cluster heads. In the fifth chapter, we proposed an energy-driven unequal clustering protocol (EDUC). Its cores are an unequal clustering algorithm and an energy-driven adaptive cluster head rotation method. The unequal size of clusters can balance the energy consumption among clusters, while the energy-driven cluster head rotation method can achieve the balance of energy consumption among nodes within a cluster. At the same time, the waste of energy is reduced. Finally, the sixth chapter summarizes this paper, and presents the future work.