DHC： Distributed, Hierarchical Clustering in Sensor Networks

• Published in: Journal of computer science and technology Vol. 26; no. 4; pp. 643 - 662
• Main Author: 马秀莉 胡海峰 李双峰 肖红梅 罗琼 杨冬青 唐世渭
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.07.2011; Springer Nature B.V; School of Electronics Engineering and Computer Science, Peking University, Beijing 100871, China; Key Laboratory of High Confidence Software Technologies(Peking University), Ministry of Education Beijing 100871, China; Key Laboratory of Machine Perception(Peking University), Ministry of Education, Beijing 100871, China%Google China, Beijing 100871, China%Department of Computer Science and Engineering, Hong Kong University of Science and Technology Clear Water Bay, Kowloon, Hong Kong, China%School of Electronics Engineering and Computer Science, Peking University, Beijing 100871, China
• Subjects: Agglomeration; Algorithms; Artificial Intelligence; Cluster analysis; Clustering; Clusters; Communication; Communications networks; Computer architecture; Computer Science; Computer simulation; Data analysis; Data mining; Data Structures and Information Theory; Datasets; Information Systems Applications (incl.Internet); Laboratories; Maintenance; Networks; Nodes; Query processing; Regular Paper; Sensors; Software Engineering; Studies; Summaries; Synthetic data; Theory of Computation; 交互式数据; 传感器节点; 分布式传感器网络; 多传感器; 大连; 层次聚类; 数据分布; 网络应用; data mining; clustering; wireless sensor networks
• ISSN: 1000-9000; 1860-4749
• DOI: 10.1007/s11390-011-1165-0

In many sensor network applications, it is essential to get the data distribution of the attribute value over the network. Such data distribution can be got through clustering, which partitions the network into contiguous regions, each of which contains sensor nodes of a range of similar readings. This paper proposes a method named Distributed, Hierarchical Clustering （DHC） for online data analysis and mining in senior networks. Different from the acquisition and aggregation of raw sensory data, DHC clusters sensor nodes based on their current data values as well as their geographical proximity, and computes a summary for each cluster. ~3arthermore, these clusters, together with their summaries, are produced in a distributed, bottom-up manner. The resulting hierarchy of clusters and their summaries facilitates interactive data exploration at multiple resolutions. It can also be used to improve the efficiency of data-centric routing and query processing in sensor networks. We also design and evaluate the maintenance mechanisms for DHC to make it be able to work on evolving data. Our simulation results on real world datasets as well as synthetic datasets show the effectiveness and efficiency of our approach.