| 摘要: |
| 随着低空经济的快速崛起，无人机集群作为低空经济的核心支撑载体，凭借其分布式协同、广域覆盖、高效作业的独特优势，已实现从实验室技术向实际生产力的跨越，在精准农业、应急救援、城市物流、国防安全等多个关键领域得到广泛应用，成为推动新质生产力发展的重要力量，其自组织网络通信链路的安全性也日益凸显。本文针对跨层协同攻击——链路垄断攻击，系统分析了其攻击机理，并提出一种动态分层防御机制。链路垄断攻击通过伪造物理层信号、污染网络层邻居列表、劫持应用层指令，实现对目标节点的隐蔽控制，具有实施门槛低、隐蔽性强等特点，现有静态防御机制难以有效应对。为此，本文构建了攻击实施的四阶段模型，并提出“三层邻居列表管理+实时信任验证+连接来源控制+异常响应优化”的综合防御方案。通过OMNeT++仿真实验表明，所提出的防御机制能够将攻击成功率降至12%，恶意节点检测率达95.8%±0.5%；消融实验进一步验证了三层邻居列表、实时信任验证、连接来源控制与异常响应优化四个模块的独立贡献与协同效应。该机制以可控的性能开销保障集群在遭受攻击时仍能维持89± 4.2%的任务完成率,为无人机自组网络的安全防护提供了一定的理论和技术参考。 |
| 关键词: 无人机自组网络 链路垄断攻击 动态防御 信任评估 邻居列表管理。 |
| DOI： |
| 投稿时间：2026-05-08修订日期：2026-07-10 |
| 基金项目: |
|
| Link Monopolization Attack and Dynamic Defense Methods for UAV Ad Hoc Networks |
|
qiuchaobang1, liweihao1, shifan2
|
| (1.National Computer System Engineering Research Institute of China;2.College of Electronic Engineeering,National University of Defense Technology) |
| Abstract: |
| With the rapid rise of the low-altitude economy, drone swarms, as the core support carriers of the low-altitude economy, have realized a leap from laboratory technology to actual productivity due to their unique advantages of distributed collaboration, wide-area coverage, and efficient operation. They have been widely applied in key areas such as precision agriculture, emergency rescue, urban logistics, and national defense security, becoming an important force in promoting the development of new types of productivity, while the security of their self-organizing network communication links has become increasingly prominent. This paper focuses on cross-layer coordinated attacks—link monopolization attacks, systematically analyzes their attack mechanisms, and proposes a dynamic hierarchical defense mechanism. Link monopolization attacks achieve covert control over target nodes by forging physical layer signals, polluting the network layer neighbor list, and hijacking application layer commands. They are characterized by low implementation thresholds and strong concealment, making existing static defense mechanisms difficult to respond effectively. To this end, this paper constructs a four-stage model for attack implementation and proposes a comprehensive defense scheme of ''three-layer neighbor list management, real-time trust verification, connection source control, and anomaly response optimization.Through OMNeT++ simulations, the proposed defense mechanism can reduce the attack success rate to 12% and achieve a malicious node detection rate of 95.8%±0.5%. Ablation experiments further verified the individual contributions and synergistic effects of the four modules: three-layer neighbor list, real-time trust verification, connection source control, and anomaly response optimization. With controllable performance overhead , this mechanism ensures that the cluster can still maintain a task completion rate of 89±4.2% under attacks, providing some theoretical and technical reference for the security protection of UAV ad hoc networks. |
| Key words: UAV ad hoc networks link monopolization attacks dynamic defense trust assessment neighbor list management. |
