Multi-agent approach for anomaly detection in automation networks

Industrial Automation systems are considered critical infrastructure as part of process and power plants. In view of the increased attacks on energy utilities, it has become a norm for the customers to be comprehended by the security requirements. Industrial Automation and Power System standards have evolved to address these requirements. An approach aligned to these standards and well integrated into the automation architecture can play a significant role in helping customer to protect the critical infrastructure from various cyber security issues. For that it really requires multi-agent based approach which protect industrial automation infrastructure at different layer. This paper proposes new solution based on multi-agent concept for the latest plant automation systems to protect customer investments through integrating security standards into the automation solutions as well as approaching more defenses in depth approaches aligning to IEC 62351 and ISA 99 standards for Industrial Control System Security.