INTELLIGENT NODE MANAGEMENT METHOD IN DISTRIBUTED TELECOMMUNICATION SYSTEMS

Abstract

The article proposes an intelligent node management method in distributed telecommunication systems based on the integration of neural network prediction, adaptive optimization, and self-organized coordination in Fog/Edge environments. The purpose of the developed method is to enhance the resilience and scalability of control processes under conditions of dynamic load variation, delays, and possible node failures. The proposed approach, implemented as the SENTRY-L (Secure Neuro-predictive Risk-aware Leader) method, provides intelligent prediction of node stability, assessment of security risks, and asynchronous transfer of coordination authority without initiating centralized election procedures. A key feature of the method is the use of a neural network to model the behavior of nodes within a cluster, allowing prediction of each node’s state based on current parameters such as bandwidth, computational resources, latency, and energy consumption. This enables a shift from reactive to proactive control, where decisions on re-electing the coordinator are made before a failure occurs. Additionally, the Security-Scoring Hub (SSH) generates a risk index and a trust matrix between nodes, integrating security directly into the coordination algorithm. Experimental modeling demonstrated that the proposed method reduces the average coordinator failure response time by 27–35% compared to classical algorithms, decreases control traffic overhead by 18–22%, and maintains decision consistency levels of 0.94–0.97 even with packet loss up to 10%.

Thus, the SENTRY-L method ensures efficient, secure, and adaptive node management in distributed telecommunication systems, combining prediction, optimization, and self-organization functions. Its implementation improves the scalability, adaptability, and resilience of next-generation Fog/Edge telecommunication networks, which is particularly relevant for applications in critical infrastructures, unmanned systems, and intelligent transport networks.