Cognitive Blockchain-Consensus Algorithm for Secure Cyber-Physical Smart Manufacturing Infrastructures

Abstract

The deployment of cyber-physical systems (CPS) in smart manufacturing has advanced industrial processes via live data analytics, automation, and intelligent decision-making. However, these interrelated systems can no longer avoid critical issues surrounding data trust, integrity, and security across heterogeneous devices and networks. The issue is to facilitate a secure, transparent, and efficient method of performing consensus in a distributed CPS environment while keeping performance and resilience in the face of cyber-attacks. This paper outlines a Cognitive Blockchain Consensus Algorithm (CBCA) for dealing with these concerns, that integrates the decentralized ledger property of blockchain with cognitive computing principles. CBCA will apply a adaptive learning model to determine consensus parameters based on real-time information in order to minimize computation overhead to reach consensus, and latency, while being able to detect threats and respond to anomalous behavior. The cognitive layer observes behavior across the network continually and makes autonomous changes to the consensus mechanism on behalf of CPS, optimizing the trust vs security tradeoff. Results from experimental simulations conducted using a smart manufacturing testbed show CBCA increasing throughput transactions by 23%, reducing consensus delay by 17%, and malicious node detection accuracy by 96% when compared to traditional Proof of Work and Proof of Stake methods.