An experimental framework for investigating hashgraph algorithm transaction speed

Abstract

Power grids around the world have experienced a growing number of malicious cyber attacks. This paper provides an overview of recent use of the Hyperledger Fabric distributed operating system to prototype use of a permissioned blockchain consensus algorithm to trust shared state estimation and control data and another effort to alter local sensor data to destroy the integrity of the shared data. The paper also provides justification for an experiment to prototype use of Babble, a peer-to-peer network plugin using the hashgraph consensus algorithm, to share the state estimation and control data through transactions recorded in a hashgraph. A key claim of the hashgraph documentation, which is unsubstantiated without a proper academic analysis, is that the algorithm is asynchronous Byzantine fault tolerance (ABFT). Also, while the Hyperledger Fabric implementation supports thousands of transactions per second, the hashgraph algorithm documentation claims orders of magnitude more. Our experiment seeks to measure the hashgraph transaction speed and determine its suitability for improving the resilience of wide area control of the smart grid. The previous resilience research of the Anomaly Detection of Cyber Physical Systems (ADCPS) team includes research into inadvertent cyber and physical failures as well as malicious attacks. We conclude with some speculations concerning the potential impact of fast, fair, and secure sharing of data across a network of blockchains potentially interfaced using hashgraph distributed ledger technology (DLT).