Home // International Journal On Advances in Intelligent Systems, volume 5, numbers 3 and 4, 2012 // View article

Fuzzy Query Propagation in Sensor Networks

Authors:
Mohamed Bakillah
Steve H.L Liang
Mir Abolfazl Mostafavi
Alexander Zipf
Jamal Jokar Arsanjani

Keywords: Fuzzy logics, information retrieval, query propagation, sensor networks

Abstract:
Query-driven information retrieval aims at supporting users to request and retrieve relevant data from sensor networks. Due to energy and capacity constraints that characterize sensor networks, information retrieval should avoid flooding the network with queries, but rather find the most efficient propagation path that maximizes the recall of relevant data while minimizing the number of sensor nodes being accessed. This is the problem of query propagation, for which numerous approaches for sensor networks have been proposed. Although, one unaddressed issue that remains is the issue of fuzziness of users’ queries and fuzziness of sensor data. When crisp criteria are used to express queries and select query recipients during propagation, some sensor nodes that are relevant can be “missed.” Therefore, this paper’s objective is to integrate a fuzzy semantic mapping mechanism, which has been published in a previous research, into a new, cluster-based fuzzy query propagation approach. The fuzzy query propagation approach avoids the overload of sensor nodes that are near the sink nodes by incorporating a first propagation step towards relevant clusters of sensors, therefore varying the sensor nodes that will have to redistribute the query, followed by an intra-cluster query propagation phase. The approach has been evaluated with a simulation and compared with a crisp version to show the impact of the consideration of fuzziness in the improvement of the recall of relevant nodes while avoiding the increase of propagation cost.

Pages: 315 to 327

Copyright: Copyright (c) to authors, 2012. Used with permission.

Publication date: December 31, 2012

Published in: journal

ISSN: 1942-2679