An Analysis of Join Processing in Sensor Networks

Wireless sensor networks have received much attention recently. Given their autonomy, flexibility and large range of functionality, they can be used as an effective and discrete means for monitoring data in many domains. Typically the network autonomy implies a limited and relatively small amount of energy for its operation. Hence, an important challenge they pose is how to process queries, i.e., manage and communicate data, in an energy-efficient manner within the network. In this paper we consider the problem of how to process join queries in a wireless sensor network. Unlike other types of queries, join queries have received little attention in the literature, despite their importance. We propose a few strategies for processing join queries, focusing on where (which sensor node(s)) to process data, and investigate their performance across several scenarios. Not surprisingly, our experiments show that no single strategy can be considered competitive for all scenarios. In order to avoid the potential high cost of using a fixed strategy for processing all queries, we develop a cost-based model that can be used to select the best join strategy for the query at hand. Our results confirm that, given a set of queries, selecting the join strategy based on the cost model is always better than using any fixed strategy for all queries.