MULTI-AGENT APPROACH TO LOAD BALANCING AND ENSURING STRUCTURAL STABILITY OF TRANSPORT TELECOMMUNICATION NETWORKS UNDER THE IMPACT OF CYBERATTACKS

Abstract

The paper addresses problem of ensuring cyber resilience of transport telecommunication networks based on multi-agent approach. It is shown that modern transport telecommunication networks are critical infrastructure objects characterized by high vulnerability to destructive cyberattacks that may cause node overloads, routing disruptions, and loss of structural connectivity. Analysis of existing approaches to transport network cybersecurity management demonstrates insufficient adaptability, autonomy, and limited ability to provide efficient load balancing and structural resilience under dynamic cyberattacks. An improved multi-agent load balancing model is proposed, which takes into account current compromise risk level of network nodes and communication channels and enables adaptive traffic redistribution through more secure routes. Proposed model is based on multi-commodity flow optimization problem with cyber risk metrics and is implemented as iterative algorithm of distributed interaction between local agents. A method of multi-agent interaction for ensuring structural resilience of transport telecommunication networks against cyberattacks is also developed. Method is based on identification of critical nodes and determination of minimum required number of backup links necessary to restore network connectivity after attacks. Proposed approach additionally provides decentralized coordination of agents during topology recovery and adaptive reconfiguration of routing structure under conditions of repeated attacks. Simulation results confirm effectiveness of proposed solutions in maintaining network operability, reducing compromise risk of routing paths, balancing traffic load, and preserving structural connectivity under destructive cyber impacts.