Agents and Secure Contracts in Cyber-Physical Systems: A Simulation

Future technologies will rely more than ever before on energy, telecommunications, and transportation networks. News of false data injections and other cyber attacks together with physical attacks pose great danger. The free and adaptable interactions between different and emerging stakeholder agents in a network provides ease of access and increased productivity but at the same time makes the network vulnerable to attacks. The vulnerabilities can be explored by using an agent-based simulation model which bypasses conventional modelling bottlenecks. One source of vulnerability is the immutability of the record of interactions between the stakeholders. We describe an agent-based network, ostensibly simulating an electricity network, where all interactions are recorded using smart contracts. We use the Mesa ABM platform for creating a community of stakeholder agents who use the Ethereum ecosystem to execute smart contracts. We show how such systems recover after being exposed to a vulnerable situation. Ours is a contribution to secure AI systems.