Clustering Algorithms For Wireless Sensor Networks

WSN (Wireless Sensor Networks) is a new measure and management network technology. It is an intelligent network application system which can implement data collection, aggregation and transmission by itself. WSN closely connect logical information world with genuine physical world, so as to implement the "computing anywhere" model. WSN has expansive application foreground in national defense and every field of national economy.WSN is intelligent and autonomous system for measure and management. It is composed of numberless tiny sensor nodes that have the ability of communicating and computing. Sensor nodes are disseminated in fields without guardianship and implement designed assignments in terms of environment by themselves. It is distributed and large-scale with resource limited. It adopts multi-hop and opposite communication mode, self-organized as well as self-adapted. WSN with large-scale data gathering must logically self-organize and deal with "any-to-one" traffic flow in topology structure. It is necessary and efficient to extract topology information that can be used to manage and control networks.Proper network topology automatically created through efficient topology control can not only improve efficiency of routing and MAC protocols , but also establish foundation for data-gathering, time-synchronizing and object-orientating, prolonging network lifecycle with saving nodes' energy. So it is especially important for sensor networks with limited energy and bandwidth to design appropriate and efficient energy-used topology control algorithm.During the research and design of my task, I red many Chinese and English literatures, and studied some previous algorithms. LEACH is a hierarchy topology control algorithm arousing more attention currently for its simplicity and efficiency, in which nodes are divided into many clusters, cluster head collecting and managing data and transmitting them to BS. The clustering mechanism adopted by LEACH accords with wireless energy-efficiency principle through multi-hop data-transmitting, and data-gathering of cluster head reduces times nodes of networks transmitting data to faraway BS which also can save energy. This papers presented two algorithms—EBAC(Energy-balanced Adaptive Clustering Algorithm) and EBAC-F (EBAC for Far-distance Communicating).EBAC made an improvement on CH (Cluster Head) selection based on LEACH, nodes' residual energy considered, and nodes' position not needed. Simulation results prove EBAC performs better than LEACH in lifetime. Moreover, due to balanced energy consumption in EBAC, all the nodes die almost at the same time (in last 20 rounds), so the reliability of sensing information in EBAC is higher than that in LEACH. EBAC-F is proposed for far-distance communicating between BS and area covered with sensor nodes. This algorithm adds the step where cluster head selecting father node based on EBAC. Father nodes collected, managed data the other cluster head transmitted and sent them to BS, reducing cluster head number communicating directly with BS. Simulation results show that EBAC-F algorithm decreases energy consumption of cluster head communicating faraway with BS, improving energy efficiency and farther prolonging network lifetime, and adapts to sensor networks with BS far away from guarding fields.