Research On Topology Control Methods For Wireless Sensor Networks

As an emerging application-oriented network, the wireless sensor networks (WSNs) has great research significance and broad application prospects with the proprieties of randomly deployed, distributed, self-organization, and etc. However, due to the limited storage capacity, computing and communication capabilities of sensor nodes, some strict requirements, such as low-power, high efficiency, are put forward on the network structure and protocols. A large number of sensor nodes, which are randomly deployed in the monitoring areas, build network topology by self-organization method, and adaptively adjust it in timely when the external conditions have changed, therefore, the methods of topology control is very important. Hence, some researches have been carried out in this paper, focusing on the topology control of clustering, scheduling in WSNs, and the main works are as follows:Combined with self-organization learning theory of artificial neural networks, competitive learning is introduced into WSNs, and a clustering algorithm is proposed based on competitive learning. Clustering is achieved by direct competition and side inhibition of winner nodes, and the conversion in six states and clustering process is described. The simulation results show that the algorithm could divide nodes into clusters which stability of number is better than LEACH, and the energy consumption is equivalent with LEACH expect the first round.Aiming at the special environment of coal mine such as long-narrow laneway, many sub-laneway and frequent accident, a new topology control algorithm suitable for coal mine is presented. The networks is initialized by election of cluster head and bridge from odd and even hops respectively, and competitive sleep and role conversion are achieved by means of competitive learning. The simulation results show that the algorithm could divide nodes into number reasonable, scale uniform, low overlap ratio clusters, and the overlap ratio is 20% when the node communication radius is 110 meters.Reference on the swarm intelligence of complex task completion by cooperation of much simple individual life, the ant colony optimization is used in event-driven WSNs, and a self-organization and self-adaptation heuristic scheduling algorithm is designed. In order to balance load and energy consumption, the monitoring data is gathered by passive method. The simulation results show that the algorithm makes the load standard deviation lower than 2% and energy consumption standard deviation lower than 1.9% of the means respectively in the first hop, and prolongs network lifetime more than 93% with the congestion degree decrease by 50% compared with MESA. Therefore, the algorithm has well proportionality of load and energy consumption.