6533b7dafe1ef96bd126eb7e
RESEARCH PRODUCT
Taxonomy of generative adversarial networks for digital immunity of Industry 4.0 systems
Svitlana GryshkoVagan TerziyanMariia Goloviankosubject
cybersecurityIndustry 4.0Computer scienceVulnerabilityneuroverkot02 engineering and technologytekoälyComputer securitycomputer.software_genreAdversarial systemImmunityTaxonomy (general)0202 electrical engineering electronic engineering information engineeringesineiden internetartificial digital immunitykyberturvallisuusGeneral Environmental ScienceFlexibility (engineering)Generative Adversarial Networksbusiness.industryMechanism (biology)020206 networking & telecommunicationsIndustry 4.0AutomationVariety (cybernetics)koneoppiminenälytekniikkaGeneral Earth and Planetary Sciences020201 artificial intelligence & image processingbusinesscomputerdescription
Abstract Industry 4.0 systems are extensively using artificial intelligence (AI) to enable smartness, automation and flexibility within variety of processes. Due to the importance of the systems, they are potential targets for attackers trying to take control over the critical processes. Attackers use various vulnerabilities of such systems including specific vulnerabilities of AI components. It is important to make sure that inappropriate adversarial content will not break the security walls and will not harm the decision logic of critical systems. We believe that the corresponding security toolset must be organized as a trainable self-protection mechanism similar to immunity. We found certain similarities between digital vs. biological immunity and we study the possibilities of Generative Adversarial Networks (GANs) to provide the basis for the digital immunity training. We suggest the taxonomy of GANs (including new architectures) suitable to simulate various aspects of the immunity for Industry 4.0 applications.
| year | journal | country | edition | language |
|---|---|---|---|---|
| 2021-01-01 | Procedia Computer Science |