Research On The Energy Efficient Clustering Algorithm Of Wireless Sensor Networks

Wireless sensor networks(WSNs) is a self-organized distributed system comprising a mass of sensor nodes that are small in size and capable of sensing, data processing, and communicating with each other. As a novel mode of information gathering technology, WSNs have made the physical world and human society closely linked. Considering the characteristics of sensor nodes such as large-scaled and randomly deployment, constraints on power and communicating ability, and frequent changes in network topology, using clustering-based hierarchical routing protocol becomes important for improving the network's quality of service. It is also fundamental in wireless sensor networks'upper layer applications.Focusing on the effectively use the energy of sensor nodes, the most efficient coverage of network and network-wide unique ID assignment, the dissertation studies on the clustering problems in wireless sensor networks. The major contribution of this dissertation is stated as follows.(1) In clustering algorithm of wireless sensor networks, to solve the problem of excessive energy consumption in the cluster heads, a residual energy and node-degree synthesized clustering algorithm ENCA is presented which is based on algorithm ACE. In cluster heads election phase of every round, it considers the residual energy and the average energy of all the nodes in each cluster, and in accordance with the node degree to optimize the selection of cluster heads in each cluster. In algorithm running phase the connection of the network is guaranteed, at the same time, it is avoided to select the node with low energy as cluster head.(2) By analyzing the impact of failure of different nodes on the coverage efficiency, an energy and coverage aware clustering algorithm, named ECCA, is proposed in this paper, which aims at improving the network coverage and extending the network lifetime. ECCA does not rely on the nodes physical location. With ECCA, the nodes in the high-density areas have high priority to be selected as cluster heads, and then the cluster head role will migrate to other nodes when its residual energy is lower than the energy threshold. The method can effectively reduce the nodes mortality. The complexity of algorithm and the choice of optimal broadcast radius are analyzed further.(3) We propose DEECIC, a distributed energy-efficient clustering algorithm that considers both coverage and ID assignment of the sensor network. DEECIC aims at clustering with the least number of cluster heads to cover the whole network and assigning an unique ID to each node based on local information. There are tradeoffs here among connectivity, energy usage, and communication latency. The simulation results show that the network performance is improved efficiently.