Scholarlink

<p>Software-Defined Networking (SDN) integrated with Digital Twins enables dynamic traffic management and centralized control, playing a crucial role in cybersecurity by enforcing policies, simulating attacks, and proactively mitigating evolving threats in real time. However, it faces challenges from stealthy attacks exploiting flow rule timeouts and cross-controller data inconsistencies. To address these limitations, Adaptive Unified Representation Analytics with Graph Entropic Neural Intelligence Ensemble is proposed for continuous monitoring, cyber-attack simulations, and proactive threat mitigation. Initially, variations in SDN controllers’ southbound and northbound interfaces create telemetry inconsistencies, causing cross-platform incompatibilities that obstruct seamless Digital Twin data aggregation and interpretation. To overcome this, the Schema-Adaptive AI-driven Generalized Ensemble Network is developed to harmonize diverse SDN telemetry to ensure unified data interpretation and a real-time cross-platform telemetry stream suitable for Digital Twin environments. Moreover, attackers exploit SDN flow rule timeouts through brief, server-shifting ephemeral C2 flows, enabling stealthy system control and undetected data exfiltration. Thus, the Adaptive Graph Entropic Intelligence Network is introduced to detect ephemeral flows and stealthy incursions by analyzing traffic entropy, adjusting flow timeouts, and uncovering hidden malicious patterns to enhance SDN resilience. Experimental results demonstrate 99.31% accuracy, 1.2s execution time, and 0.68 loss, ensuring real-time anomaly detection and proactive threat response. This significantly strengthens SDN security against stealthy cyber incursions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Digital Twin-SDN security with adaptive graph entropic intelligence and cross-platform schema harmonization

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月29日

查看全文
  • Gaurav Prasad,
  • Devi Prasad Sharma

摘要

Software-Defined Networking (SDN) integrated with Digital Twins enables dynamic traffic management and centralized control, playing a crucial role in cybersecurity by enforcing policies, simulating attacks, and proactively mitigating evolving threats in real time. However, it faces challenges from stealthy attacks exploiting flow rule timeouts and cross-controller data inconsistencies. To address these limitations, Adaptive Unified Representation Analytics with Graph Entropic Neural Intelligence Ensemble is proposed for continuous monitoring, cyber-attack simulations, and proactive threat mitigation. Initially, variations in SDN controllers’ southbound and northbound interfaces create telemetry inconsistencies, causing cross-platform incompatibilities that obstruct seamless Digital Twin data aggregation and interpretation. To overcome this, the Schema-Adaptive AI-driven Generalized Ensemble Network is developed to harmonize diverse SDN telemetry to ensure unified data interpretation and a real-time cross-platform telemetry stream suitable for Digital Twin environments. Moreover, attackers exploit SDN flow rule timeouts through brief, server-shifting ephemeral C2 flows, enabling stealthy system control and undetected data exfiltration. Thus, the Adaptive Graph Entropic Intelligence Network is introduced to detect ephemeral flows and stealthy incursions by analyzing traffic entropy, adjusting flow timeouts, and uncovering hidden malicious patterns to enhance SDN resilience. Experimental results demonstrate 99.31% accuracy, 1.2s execution time, and 0.68 loss, ensuring real-time anomaly detection and proactive threat response. This significantly strengthens SDN security against stealthy cyber incursions.