Optimized data-available storage for energy-limited wireless sensor networks

Environmental monitoring with wireless sensor networks (WSNs) is based on an infrastructure with unreliable communication channels and nodes. WSNs, also, have a limited lifetime due to the fixed energy budget. Nevertheless, the user of a data-storing WSN often needs high availability of the measured data. But unfortunately, there is no conceptional framework to articulate those QoS requirements nor a design-guideline that leads to a WSN implementation guaranteeing these requirements. In this paper, we address these problems. We present an approach that identifies the organization of the measured data within the WSN giving minimal availability guarantees as demanded by the end-user at minimal energy consumption. We rely on sensor data replication in order to implement query-specific availability guarantees in an energy-optimized manner. We exemplary show the feasibility of our approach through an adaptation of the well- known WSN experiment from Great Duck Island.