STTR Phase II: Performant Distributed Ledgers for Cybersecurity Applications

The broader/commercial impact of this STTR Phase II project is to allow distributed ledgers to naturally scale as their workloads do, reducing the barrier for companies to adopt and operate the technology over the long-term. A distributed ledger is a database shared by multiple participants in which each participant or node maintains and updates a synchronized copy of the data. Distributed ledgers allow members to securely verify, execute, and record their own transactions without relying on a central intermediary and provide benefits such as immutability, decentralization of data, and high availability. Distributed ledger technology (DLT) is quickly growing in adoption both in the US and globally. As is true with any emerging technology, distributed ledgers need solutions to support scale, particularly in the context of ledger growth, fast search, and tractable cost of deployment over time. Further the project will advance private, permissioned blockchains to become viable for production-scale, enterprise use cases spanning sectors like cybersecurity and finance, which both have uncompromising real-time, high transaction throughput constraints. This STTR Phase II will tackle two key technical objectives: 1) Pruning in enterprise-type DLT systems and 2) Dynamic scaling of deployed DLT platforms. This research will be performed in the context of a widely adopted and open-source DLT technology but will study the implications of both pruning and scaling on factors including security, privacy, consensus, and implementation complexity on distributed ledger technology (DLT) more broadly. To accomplish these objectives, the project will rely on findings from experiments executed and methods developed during the Phase I SBIR research. These include the use of realistic, highly scalable load-testing infrastructure to simulate and benchmark thousands of simultaneous clients. A major research objective of this Phase II work is to develop algorithms for context-aware transaction compression pruning. This approach to DLT pruning is well suited for enterprise applications where the transaction itself may store a variable amount of data. This pruning mechanism would substantially advance the state-of-the-art in transaction and state pruning, as it would not require re-syncing of nodes, nor suffer information loss locally, nor damage the ability to reach consensus, and will be optimized for fast search. Further, this research will examine different classes of applications and discuss which of these could benefit from Corsha’s novel pruning technology. 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.