Перегляд за Автор "Stopochkina, I. V."

Зараз показуємо 1 - 3 з 3
  • Ескіз
    ДокументВідкритий доступ
    A Multi-Agent Approach to Detecting Cyber-Physical Attacks in a Turbine Control System Based on the HAI 23.05 Dataset
    (КПІ ім. Ігоря Сікорського, 2026) Holubieva, I. S.; Stopochkina, I. V.
    Industrial control systems in energy facilities are vulnerable to cyber-physical attacks. Such attacks can affect both digital control components and physical process variables. This work addresses attack detection in a turbine control system using the HAI 23.05 dataset. A multi-agent architecture is proposed for analyzing different attack types and signal groups. The agents focus on speed setpoint manipulation, controller output modification, trip protection parameter changes, and cross-process consistency violations. The architecture defines data transfer from the experimental stand to the analytical layer through SCADA/Historian, OPC-UA, and industrial Ethernet. The main result is the decomposition of anomaly detection into specialized agents, supported by attack-scenario analysis and a preliminary Isolation Forest baseline.
  • Ескіз
    ДокументВідкритий доступ
    Model and Method of Policy-Driven Backdoor Detection on Endpoints with Secure Signature Distribution and Optimized Scan Execution
    (КПІ ім. Ігоря Сікорського, 2026) Protsyshyn, I. I.; Stopochkina, I. V.
    This paper proposes a policy-driven model for backdoor detection on endpoints, in which signature-based scanning is combined with scan execution policies and secure rule-set updates. The practical implementation of the approach is oriented toward a prototype that includes an endpoint agent, a trusted local cache of signature sets, scanning profiles, and a YARA-compatible engine. A priority function is proposed that takes into account whether an object belongs to autorun locations, whether it has been modified recently, whether it is located in a privileged area, and whether it lacks a digital signature. Secure signature updates include verification of the signature, hash, and version number of the rule set. The result of the work is a structural diagram of the prototype, a fragment of a YARA rule, and an experimental evaluation scheme focused on comparing full and priority-based scanning.
  • Ескіз
    ДокументВідкритий доступ
    Practical Validation of a Scenario-based Risk Engine for Semi-quantitative Assessment of Hybrid Risks to Critical Mobile Operator Services
    (КПІ ім. Ігоря Сікорського, 2026) Seriakov, V. L.; Stopochkina, I. V.
    The paper presents the practical validation of a scenario-based decision-support risk engine for semi-quantitative assessment of hybrid risks to critical services of a mobile operator. Hybrid risks are treated as scenarios in which cyber, physical, transport, power-related, operational, and organisational dependencies jointly affect service availability, integrity, or manageability. The proposed approach links structural scoring, expert assessment, control-aware residual-risk recalculation, heatmap analysis, disagreement analysis, and what-if modelling. The validation case includes 23 elements, 22 scenarios, 8 expert roles, 176 expert scores, and 25 controls. The results show that the prototype can prioritise critical scenarios, explain the origin of risk scores, identify expert-disagreement cases, and support residual-risk reduction decisions.