An Energy Efficient Design of Multi-resolution Storage for Ubiquitous Data Management

Ubiquitous computing enables dense sensing of the environment, offering unprecedented opportunities for observing the physical world. In many scientific applications in dense sensor arrays, efficient support for multi-resolution storage and drill-down queries is essential. We argue that technology trends in embedded platforms and flash memories call for rethinking multi-resolution storage designing. We articulate a vision of a storage-centric sensor network where sensor nodes will be equipped with high-capacity and energy-efficient local flash storage. The data are pulled or pushed from remote sensors by proxies, which are assumed to be less resource and energy constrained than remote sensors. We then describe how COSA enables this vision through some innovative features including energy-efficient use of flash memory, multi-resolution summarization and aging, query processing, and caching.