Diagnostic behavior analysis of profuse data intrusions in cyber physical systems using adversarial learning...

Area of Science:

• Computer Science
• Cybersecurity
• Data Science

Background:

• Cyber Physical Systems (CPS) are increasingly vulnerable to data intrusions due to expanding network topologies.
• Data integrity is crucial, as compromised data can lead to significant harm to users and systems.
• Existing data management strategies struggle to effectively eradicate invasion-associated data.

Purpose of the Study:

• To propose a novel Cyber Physical Systems (CPS) framework for mitigating data intrusions.
• To develop an analytical framework capable of distinguishing true from false data samples and assessing failure rates.
• To minimize data loss and eliminate incursions through an integrated learning technique.

Main Methods:

• Construction of a distinctive system model with an analytical framework for CPS.
• Integration of a learning technique employing generators and discriminators to reduce data loss.
• Real-time evaluation and simulations of the proposed integrated framework.

Main Results:

• The proposed CPS framework demonstrated significantly reduced failure rates and data losses in simulations.
• The learning-based approach achieved minimal data loss (1%) and high efficacy (97%).
• Simulated results outperformed existing methodologies lacking learning components.

Conclusions:

• The developed CPS framework effectively mitigates intrusions with minimal data loss.
• The framework is suitable for real-time implementation without compromising data integrity.
• The learning-based approach offers a superior solution for securing Cyber Physical Systems.