Collaborative Research: CNS Core: Medium: Secure, Reliable, and Efficient Long-Term Storage

The world produces more digital data annually that’s growing exponentially. Storing data for decades is challenging and securing it for so long is even more challenging. This project develops secure, efficient, long-term archival storage systems for digital information that endures beyond a single human lifetime.

The project uses a long-term security model that is safe against faster computers and even quantum computers and can protect against malicious "insiders" who abuse their access to secretly steal data slowly over years. Lastly, the project explores how to maintain long-term data securely and efficiently, even if individual storage providers cease to exist.

This project develops techniques to store archival data securely and reliably for a long-time using information theoretic security and combinatorial security. The project explores techniques to allow data to survive errors ranging from corruptions of a few bytes to large-scale cloud failures. It then explores the trade-off between additional storage requirements, data security and reliability, and system performance.

Empirical evaluation of a prototype system provides insights into real-world issues in implementing these techniques over the project’s lifespan, and a simulator embodying these techniques allows projection of the techniques’ effectiveness over much longer time frames.

This project fosters collaborations across systems, theory, and security researchers to develop practical techniques to both secure data for many years and ensure that the data endures unchanged. Storage must meet both criteria for a society based on digital data to rely on it. All source code for a prototype system and simulator is maintained and released publicly.

Material from this project is integrated into graduate level courses and a new “Secure Storage Systems” course. The project recruits and co-advises several under/graduate students, with a focus on female and Hispanic students, both traditionally underrepresented in computer systems research.

The project's artifacts—software, source code, data sets, secure archive simulator, traces, and results—are all embodied in a system called "SecArch: Secure Archives". Results will be disseminated using peer-reviewed publications and arxiv.org. All artifacts will be made public through the project Website: https://www.filesystems.org/secarch. The project plans to maintain the site for at least ten years following the end of the project.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.