An adaptive zone-based storage architecture for wireless sensor networks

In a large-scale sensor network, the amount of data collected increases rapidly over time. Efficient data storage and retrieval in sensor networks is therefore critical to enhancing network lifetime. In this paper, we propose an adaptive zone-based storage architecture for sensor networks. Our design consists of three major components: 1) a distributed clustering algorithm that operates on a novel concept of node contribution potential, allowing the network to be clustered based on desired performance, 2) a local storage policy to reduce cross-network transmission which is resilient to node and network failures, and 3) an overlay of zone leaders to distribute queries and data efficiently. In cases where the query sources are limited to a number of fixed points in the network, we propose a query source adaptation policy to reduce network stress on overlay links. The strengths of our architecture include the ability to customize the network's desired performance at run time, its low protocol overhead, scalability and resilience to network failures.