MilTAP: A Tunnel-Aware Authentication and AI-Driven Policy Enforcement Protocol for Military VPN Networks

Abstract

In modern military networks, Virtual Private Networks (VPNs) are essential for safeguarding sensitive information across distributed operational and command facilities. However, conventional authentication frameworks and static firewall policies often lack the adaptability to secure dynamic VPN tunnels, exposing systems to sophisticated adversarial threats. This paper presents MilTAP, a Tunnel-Aware Authentication and AI-Driven Policy Enforcement Protocol designed for the stringent security requirements of military VPN infrastructures. MilTAP integrates robust cryptographic methods for mutual authentication with continuous tunnel state assessment to ensure resilient communications. An embedded AI-based anomaly detection module analyzes real-time telemetry, evaluating tunnel health and session traffic to differentiate legitimate activity from potential attacks. When anomalies are detected, the protocol dynamically adjusts firewall rules to limit or revoke access, mitigating threats such as session hijacking and credential replay. Formal security analysis based on the Real-or-Random model confirms session key indistinguishability against probabilistic adversaries. Verification with AVISPA and ProVerif further demonstrates resilience to diverse attack vectors while preserving operational performance. Experimental results indicate MilTAP achieves low computational overhead and rapid response times, supporting its suitability for mission-critical defense environments.