An optimal distribution of data reduction in sensor networks with hierarchical caching

Abstract

Reducing data transmission to conserve energy and provide cost-efficient query answers in sensor networks is the topic of this paper. We consider a sensor network logically organized into a tree structure. The sensors at the lowest level cache raw data, and data is stored with greater degree of compression as we move up the tree. Thus a query answered using data at a lower level has less error but requires more energy. We have proposed a model for accounting the cost of resources consumed and the error in the result obtained. We have formulated an optimization problem for the trade-off between the error cost and the resource cost of answering queries. Its solution enables us to determine the optimal distribution of data reduction at each level. We have presented numerical solutions for some sample data, illustrating the practicality of the scheme. © IFIP International Federation for Information Processing 2007.