Abstract
The article presents a comprehensive and in-depth review of multi-vector cyberattacks implemented in cyber-physical systems (CPS). The study provides a detailed analysis of the key architectural prerequisites of CPS, threat models at the intersection of information technology and operational technology (IT/OT), as well as the mechanisms for coordinating cyber and physical impacts. The main types of attacks are systematized, including false data injection (FDI/FDIA), malware targeting industrial control systems and PLCs, supply-chain attacks, ransomware, denial-of-service (DoS) attacks, insider threats, and direct physical interference. Their interconnection and amplification effects within a unified compromise scenario are demonstrated. Based on open scientific studies and industry reports, the paper analyzes notable cases such as Stuxnet, cyberattacks on Ukraine’s energy infrastructure, CrashOverride/Industroyer, TRITON/Trisis, and ransomware incidents in industrial environments. In addition, a risk-oriented matrix based on the principle «Attack Type – CPS Layer» is proposed. A practical anomaly-detection pipeline using IT/OT event correlation and machine-learning methods is described. The work includes tables, diagrams, and graphs illustrating CPS architecture, the lifecycle of multivector operations, and engineering security measures aligned with the standards ST RK ISO/IEC 27001, NIST, ISA/IEC 62443, and MITRE ATT&CK for ICS.