A GenAI-Driven Adaptive Cybersecurity Mesh for Real-Time Threat Detection in Intelligent Communication Systems

Abstract

Intelligent communication systems integrating Internet of Things (IoT), cyber-physical infrastructures, edge computing, and next-generation communication protocols have significantly expanded the attack surface of modern digital ecosystems. Traditional intrusion detection systems (IDS) remain predominantly centralised, signature-based, and insufficiently adaptive to evolving multi-vector and zero-day threats. This paper proposes a GenAI-driven adaptive cybersecurity mesh architecture designed for real-time threat detection in distributed intelligent communication environments. The proposed framework integrates zero-trust security principles with a distributed mesh of edge security nodes coordinated through a policy orchestration layer. A generative AI-based adaptive threat modelling engine continuously synthesises contextual attack patterns and enhances anomaly detection across network, application, and behavioural layers. A formal cross-layer risk-scoring model fuses heterogeneous security signals to generate dynamic threat-confidence indices. The system is evaluated in a simulated intelligent communication environment comprising heterogeneous nodes, mixed legitimate traffic, and multiple attack scenarios, including DDoS, man-in-the-middle, and protocol-exploitation attacks. Experimental results demonstrate improved detection accuracy, reduced false positives, and lower response latency compared to baseline signature-based and centralised ML-based IDS models. The proposed architecture offers a scalable and adaptive security paradigm suitable for next-generation intelligent communication infrastructures.