The proposed project centers on data storage as a target application for research innovations that span theory and practice. The objective of the project is to improve the key features of merit in data storage: storage density, data reliability, access performance and security. These features and the tradeoffs between them will be treated by introducing and advancing theoretical frameworks in various information sciences. To prove and measure these improvements, systems research provides important complementary tools. The project will therefore carry out a joint theory and systems research to solve the most critical problems in data storage.
The project is timely in the sense that separate efforts in theory and in systems are no longer effective to deliver the desired performance. In all known storage technologies, scaling in density and in access speed can no longer be achieved by physical means only. Rather, clever methods now need to be developed for the placement and low-level representation of data within the device. With harsher design conditions, data placement has to be well aware of the low-level media properties, and data representation must be designed with the device access properties in mind. These new design dependencies require a joint research effort in theory and in systems to find new analytic constructs and understand their performance in real storage environments.
Previous contributions by the proposing researcher while at academic and industrial research institutes show the great promise of this line of research. Different ubiquitous and emerging storage applications have already benefited from this research: solid-state storage, next-generation magnetic storage and storage systems.
Call for proposal
See other projects for this call