A Clustering Method of Combining Grid and Genetic Algorithm in Wireless Sensor Networks
This paper presents a clustering method of combining grid and genetic algorithm (GA) based on grid and global optimization in wireless sensor networks (WSN). The algorithm first partitions grid based on node’s location, then computes clustering center of
- PDF / 258,704 Bytes
- 7 Pages / 439.37 x 666.142 pts Page_size
- 63 Downloads / 162 Views
A Clustering Method of Combining Grid and Genetic Algorithm in Wireless Sensor Networks Jun Zeng
Abstract This paper presents a clustering method of combining grid and genetic algorithm (GA) based on grid and global optimization in wireless sensor networks (WSN). The algorithm first partitions grid based on node’s location, then computes clustering center of grid using membership degree of GA, and then introduces dimensionality reduction pretreatment of the high-dimensional samples mapping into the two-dimensional space and optimum maintaining strategy. Simulation results show that the method in this paper can reduce iteration times and clustering accuracy is higher. Keywords Wireless sensor networks (WSN) Clustering
Grid Genetic algorithm (GA)
95.1 Introduction Wireless sensor networks (WSN) are made up of a large number of wireless sensors nodes which can compute, perceive, and wirelessly communicate through network of self-organization, it can independently complete monitoring, target detection, tracking, and other tasks according to the environment [1]. WSN had become a new computing platform, each node has sensing within environment, data processing, and wireless communication ability. WSN are data-centric network, if we search some data needs in a number of perceived data, the clustering technology is one of key technologies. However, in wireless sensor network,
J. Zeng (&) Yangtze Normal University, Fuling, Chongqing 408100, China e-mail: [email protected]
Z. Zhong (ed.), Proceedings of the International Conference on Information Engineering and Applications (IEA) 2012, Lecture Notes in Electrical Engineering 218, DOI: 10.1007/978-1-4471-4847-0_95, Springer-Verlag London 2013
773
774
J. Zeng
sensor nodes have low computing power, small storage capacity, slow communication, and so on, which makes wireless sensor network itself cannot complexly and timely process a large number of sensor data within a short time. So how to make full use of limited energy and storage capacity to design efficient querying and processing technology is a hot research currently.
95.2 Clustering Based on Grid Grid technology is a technology of distributed computing in recent years; grid can connect high-speed computers, large databases, and storage devices together, and provide users a unified grid service. Using grid for data fusion, it can play advantages of a huge amount computing and storage resources in grid, and process, analysis, and store to collect a large amounts of data in the wireless sensor network, and make data more fully treated in WSN. Data fusion using grid in WSN also allows integration of multiple wireless sensor network systems, shields heterogeneity of data in WSN, and data from a wireless sensor network can be used by multiple grids to improve utilization of data. It simplified operation of user in WSN, and provided a unified interface of data. In grid, we can obtain new knowledge from technology of data mining, data fusion, distributed database, and so on. Grid computing in which indepen
Data Loading...
