Operational Mechanisms and Environment Models of the Multi-Agent System

Khorolska, Karyna та Bebeshko, Bohdan (2025) Operational Mechanisms and Environment Models of the Multi-Agent System Колективна (три і більше авторів). Estonia, Scientific Center of Innovative Research.

Текст K_Khorolska_B_Bebeshko_ITSC_2025.pdf Download (1MB)

Анотація

Enterprise information systems are exposed to persistent and fast-changing cyber threats, while their distributed structure makes single-point detection and response increasingly fragile. This chapter proposes an operational model for a multi-agent system (MAS) that distributes sensing, analysis, and counteraction across interconnected environments, including workstations, network segments, subnetworks, servers, and perimeter routers. Agent interactions are adaptively driven by threat levels inferred by neural networks, and communication is refined through load balancing and delay-minimisation principles to avoid congestion. The objective is to specify operational mechanisms and environment models that enable scalable, resilient, and intelligent cyber defence in enterprise infrastructures, with coordinated decisions that remain robust under advanced and persistent attacks. The MAS is defined as a set of agents and environments, with formalised interaction intensity, distribution constraints, and communication delays to control overhead and preserve responsiveness. Threat evaluation is grounded in neural outputs with adaptive thresholds, complemented by fuzzy logic and trust modelling to cope with uncertainty and unreliable signals. Collective decision-making is operationalised through voting procedures, including Condorcet and Nanson methods, with trust-aware weighting to mitigate cyclical preferences. The study delivers a hierarchical MAS architecture with multi-level message propagation and modular subsystems for sensing, analysis, trust evaluation, and secure communication, coordinated by a central control component. Simulation in OMNET++ under realistic scenarios, including DDoS, indicates reduced response time, improved accuracy, and productive cooperation among specialised agents. A formally structured, trust-aware MAS can strengthen enterprise cyber defence by combining distributed monitoring with coordinated, adaptive responses, while maintaining communication efficiency and operational scalability across heterogeneous infrastructure layers.

Тип елементу :	Монографія (Колективна (три і більше авторів))
Ключові слова:	multi-agent system; enterprise cybersecurity; environment modelling; attack detection; incident response; adaptive thresholds; neural networks; fuzzy logic; trust modelling; Condorcet voting; Nanson method; OMNET simulation
Типологія:	Монографії > Видані в іноземному видавництві мовами ОЕСР/ЄС
Підрозділи:	Факультет інформаційних технологій та математики > Кафедра інформаційної та кібернетичної безпеки ім. професора Володимира Бурячка
Користувач, що депонує:	Павло Миколайович Складанний
Дата внесення:	13 Лют 2026 10:23
Останні зміни:	13 Лют 2026 10:23
URI:	https://elibrary.kubg.edu.ua/id/eprint/56410

Actions (login required)

Перегляд елементу