Node Wake-Up via OVSF-Coded Bloom Filters in Wireless Sensor Networks

Abstract

Interest dissemination in constrained environments such as wireless sensor networks utilizes Bloom filters commonly. A Bloom filter is a probabilistic data structure of fixed length, which can be used to encode the set of sensor nodes to be awake. In this way an application can disseminate interest in specific sensor nodes by broadcasting the Bloom filter throughout the complete wireless sensor network. The probabilistic nature of a Bloom filter induces false positives, that is some sensor nodes will be awake without the application having interest in their sensor values. As the interest is often depending on location such as in adaptive sampling applications, we present a novel method to encode both interest and possible location of information into one probabilistic data structure simultaneously. While our algorithm is able to encode any kind of tree-structured information into a fixed length bit array we exemplify its use through a wireless sensor network. In comparison to traditional Bloom encoding techniques we are able to reduce the overall number of false positives and furthermore reduce the average distance of false positives from the next true positive of the same interest. In our example this helps to reduce the overall energy consumption of the sensor network by only requesting sensor nodes that are likely to store the requested information.