Network Comprehension by Clustering Streaming Sensors

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 2010

Autores

Pedro Rodrigues (Autor) FCUP; João Gama (Autor) FCUP; João Araújo (Autor) FCUP; Luís Lopes (Autor) FCUP

Ata de Conferência Internacional

Título: Proceedings of the Fourth International Workshop on Knowledge Discovery from Sensor Data (Sensor-KDD 2010)

Páginas: 35-44

4th International Workshop on Knowledge Discovery from Sensor Data (Sensor-KDD 2010)

Washington, DC, 25 - 28 July, 2010

Classificação Científica

FOS: Ciências exactas e naturais > Ciências da computação e da informação

CORDIS: Ciências Físicas > Ciência de computadores

Outras Informações

Abstract (EN): Sensor network comprehension tries to extract information about global interaction between sensors by looking at the data they produce. When no other information is available to extract, usual knowledge discovery approaches are based on unsupervised techniques. However, if these techniques require data to be gathered centrally, communication and storage requirements are often unbounded. The goal of this paper is to discuss sensor network comprehension techniques, presenting a local algorithm to compute clustering of sensors at each node, using only neighbors' centroids, as an approximation of the global clustering of streaming sensors computed by a centralized process. The clustering algorithm is based on the moving average of each node's data over time: the moving average of each node is approximated using memoryless fading average; clustering is based on the furthest point algorithm applied to the centroids computed by the node's direct neighbors. The algorithm was evaluated on a state-of-the-art sensor network simulator, measuring the agreement bewteen local and global clustering. Results show a high level of agreement between each node's clustering denitions and the global clustering denition, with special emphasis on separability agreement. Overall, local approaches are able to keep a good approximation of the global clustering, improving the ability to keep global network comprehension at each sensor node, with increased privacy, and decreased communication and computation load in the network.

Idioma: Inglês

Tipo (Avaliação Docente): Científica

Notas: Disponível em: http://www.ornl.gov/sci/knowledgediscovery/SensorKDD-2010/SensorKDD'10_Proceedings.pdf