RESTORE: A Real-Time Event Correlation and Storage Service for Sensor Networks

This paper describes the design of RESTORE, which is a framework for providing in-network event correlation and storage service for sensor environments. RESTORE uses a data-centric approach in which it partitions a sensor network into zones and maps every event to a zone. The sensor nodes in a zone make use of their cooperative storage resources and redundancy to improve information availability and energy efficiency. RESTORE is useful for temporarily buffering information in sensor environments that have intermittent connectivity to a base station. RESTORE may also be used as the underlying framework for an event notification service that publishes events directly from a sensor network to the subscribers. The contributions of this paper include a) an event taxonomy for correlating sensor events within the network, and b) mechanisms to organize information using the collaborative resources, in order to enable real-time event correlation within a sensor network. We also present preliminary simulation results that examine how the collaboration within a zone in RESTORE impacts the energy consumption, information availability, and message overhead.