WindGuard: An autonomic self-defense agent

Cyberattacks targeting offshore wind power plant (WPP) industrial networks are increasing in both frequency and sophistication, creating serious risks for operational continuity. Such attacks can trigger unplanned shutdowns of wind turbine generators, disrupt SCADA operations, delay fault detection and maintenance, and in severe cases, cause long-term damage to critical assets.This dataset and code package, termed WindGuard, supports the development and evaluation of an autonomic self-defense agent based on OpenFlow Random Host Mutation (OF-RHM). The agent dynamically reshapes the network attack surface to increase the complexity and cost of intrusions, enabling it to anticipate, withstand, recover from, and adapt to cyber threats. Included are:Network topology definitions and emulation scenarios for Mininet+ONOS,Implementation of the OF-RHM agent,Experiment traces covering reconnaissance, DDoS, and stealthy crossfire attacks, andLong-term resilience modeling scripts using Generalized Stochastic Petri Nets (GSPNs).Testbed experiments demonstrate that WindGuard successfully mitigates multiple attack types with up to 80% containment, while resilience modeling yields indices ranging from 76.4% to 99.7% across various scenarios. These resources provide a reproducible foundation for advancing cyber-resilient, software-defined industrial networks in offshore WPP environments.