大语言模型增强的抗灰洞攻击海域无人机路由算法

李杰铃; 肖亮; 王鹏程; 雷妍; 陈乔鑫; 王成耀

doi:10.12263/DZXB.20250878

您当前的位置：

首页 >

文章列表页 >

大语言模型增强的抗灰洞攻击海域无人机路由算法

学术论文 | 更新时间：2026-04-24

- 大语言模型增强的抗灰洞攻击海域无人机路由算法
- LLM-Enhanced Maritime UAV Routing Algorithm Against Gray-Hole Attacks
- 电子学报 2025年53卷第12期页码：4474-4484
- 作者机构：
  
  厦门大学信息学院，福建厦门 361102
- 作者简介：
  
  [ "李杰铃男，1995年1月出生于福建省泉州市.现为厦门大学信息学院博士研究生.主要研究方向为网络安全、无人机组网等.E-mail: 18759923858@163.com" ]
  [ "肖亮女，1980年3月出生于河北省石家庄市.现为厦门大学信息学院教授、博士研究生导师.主要研究方向为无线通信、网络安全等.E-mail: lxiao@xmu.edu.cn" ]
  [ "王鹏程男，2002年8月出生于江苏省苏州市.现为厦门大学信息学院硕士研究生.主要研究方向为无人机组网、无线安全认证.E-mail: amat_wpc@163.com" ]
  [ "雷妍女，2001年5月出生于广东省深圳市.现为厦门大学信息学院博士研究生.主要研究方向为无人机组网、网络安全等.E-mail: 23320250157879@stu.xmu.edu.cn" ]
  [ "陈乔鑫男，1998年7月出生于福建省泉州市.现为厦门大学信息学院博士研究生.主要研究方向为无线安全、智能安全检测等.E-mail: fjqiaoxin_chen@163.com" ]
  [ "王成耀男，2002年7月出生于广东省揭阳市.现为厦门大学信息学院硕士研究生.主要研究方向为无人机组网、无线抗干扰等.E-mail: 19876818080@163.com" ]
- 基金信息：
  
  国家自然科学基金(U21A20444);中央高校基本科研业务费专项资金(20720250036);国家重点研发计划(2023YFB3107603)
- DOI：10.12263/DZXB.20250878
  中图分类号： TP393;
- 收稿：2025-11-25，
  
  录用：2025-12-18，
  
  纸质出版：2025-12-25
- 稿件说明：
移动端阅览
李杰铃, 肖亮, 王鹏程, 等. 大语言模型增强的抗灰洞攻击海域无人机路由算法[J]. 电子学报, 2025, 53(12): 4474-4484.

LI Jie-ling, XIAO Liang, WANG Peng-cheng, et al. LLM-Enhanced Maritime UAV Routing Algorithm Against Gray-Hole Attacks[J]. Acta Electronica Sinica, 2025, 53(12): 4474-4484.
李杰铃, 肖亮, 王鹏程, 等. 大语言模型增强的抗灰洞攻击海域无人机路由算法[J]. 电子学报, 2025, 53(12): 4474-4484. DOI：10.12263/DZXB.20250878

LI Jie-ling, XIAO Liang, WANG Peng-cheng, et al. LLM-Enhanced Maritime UAV Routing Algorithm Against Gray-Hole Attacks[J]. Acta Electronica Sinica, 2025, 53(12): 4474-4484. DOI：10.12263/DZXB.20250878

摘要

无人机路由实现图像、音频和位置等多模态数据至部署大语言模型的船载目的节点的高效传输，支撑目标搜索等推断任务，适用于环境监测和搜索救援等海域业务.然而，在恶劣海域信道条件下，无人机网络拓扑快速变化，路由稳定性显著下降.同时，灰洞攻击选择性丢弃数据包，导致感知数据传输的丢包概率和时延大幅增加，甚至引发推断失败.为此，本文提出大语言模型增强的抗灰洞攻击海域无人机路由算法，根据大语言模型所推断出的环境特征以及相邻无人机成功转发的数据包数，建立路由信任体系，采用强化学习优化下一跳无人机和发射功率.面向海域业务服务质量需求，结合相邻无人机信任度，设计路由策略分布函数，适配无人机群的网络拓扑和信道变化，快速恢复中断路由.针对海上节点分布稀疏及信道快速变化导致的反馈丢失问题，在路由经验回放中引入反馈恢复机制，提升路由稳定性.搭建海域无人机路由系统，在船载目的节点部署参数量为70亿的大模型LLaVA-1.5，以图像和相邻无人机位置等信息为输入，识别环境特征并将结果反馈给无人机，增强路由策略优化.基于厦门欧厝海域实测信道数据，构建30架无人机防御具有不同丢包概率灰洞攻击的结果表明，所提算法可提升72.8%的数据包到达率，降低75.1%的端到端时延和64.7%的能耗，支撑海域大语言模型任务.

Abstract

Unmanned aerial vehicle (UAV) routing enables the efficient transmission of multimodal data such as images

audio and location to the shipborne destination node equipped with a large language model (LLM) to support inference tasks including target search

which are applicable to maritime applications such as environmental monitoring and search and rescue. However

UAV network topologies change rapidly under harsh maritime channel conditions

resulting in significant degradation of routing stability. Meanwhile

gray-hole attacks selectively discard packets

leading to substantial increases in packet loss rate and transmission latency

and even causing inference failures. To address these challenges

this paper proposes an LLM-enhanced maritime UAV routing algorithm against gray-hole attacks that exploits the environment feature inferred by the LLM and the number of packets successfully forwarded by neighboring UAVs to construct a routing trust framework and applies reinforcement learning to jointly optimize the next hop UAV and the transmit power. The routing policy distribution function is formulated based on the quality-of-service requirements and the trust levels of the neighboring UAV

enabling rapid self-healing in response to dynamic network topologies and channel variations. To address feedback loss caused by sparse node distribution and rapidly varying channels in maritime environments

a feedback recovery mechanism is incorporated into routing experience replay to enhance routing stability. We develop a maritime UAV routing system

with the shipborne as the destination hosting a 7-billion-parameter LLaVA-1.5 model. Taking the captured images and one-hop neighbor information such as location as input

this model infers environment features and feeds the results back to UAVs to enhance the routing policy optimization. Based on measured channel data from the Oucuo sea area in Xiamen

a simulation scenario is constructed with 30 UAVs under gray-hole attacks with different packet loss probabilities. The results show that the proposed algorithm improves 72.8% packet delivery ratio

reduces 75.1% end-to-end latency and 64.7% energy consumption

and effectively supports LLM-driven maritime applications.

关键词

Keywords

references

LIU B C , ZHANG W K , CHEN W H , et al . Online computation offloading and traffic routing for UAV swarms in edge-cloud computing [J ] . IEEE Transactions on Vehicular Technology , 2020 , 69 ( 8 ): 8777 - 8791 .

YANG T T , JIANG Z , SUN R J , et al . Maritime search and rescue based on group mobile computing for unmanned aerial vehicles and unmanned surface vehicles [J ] . IEEE Transactions on Industrial Informatics , 2020 , 16 ( 12 ): 7700 - 7708 .

文鹏 , 叶苗 , 王勇 , 等 . SDWN中基于多智能体图强化学习的多对多通信路由方法 [J ] . 电子学报 , 2025 , 53 ( 6 ): 1885 - 1905 .

WEN P , YE M , WANG Y , et al . A multi-agent graph reinforcement learning method for many-to-many communication routing in SDWN [J ] . Acta Electronica Sinica , 2025 , 53 ( 6 ): 1885 - 1905 . (in Chinese)

孙鹏浩 , 兰巨龙 , 申涓 , 等 . 一种基于深度增强学习的智能路由技术 [J ] . 电子学报 , 2020 , 48 ( 11 ): 2170 - 2177 .

SUN P H , LAN J L , SHEN J , et al . An intelligent routing technology based on deep reinforcement learning [J ] . Acta Electronica Sinica , 2020 , 48 ( 11 ): 2170 - 2177 . (in Chinese)

WANG Z L , YAO H P , MAI T L , et al . Learning to routing in UAV swarm network: A multi-agent reinforcement learning approach [J ] . IEEE Transactions on Vehicular Technology , 2023 , 72 ( 5 ): 6611 - 6624 .

ZHANG X C , ZHAO H T , XIONG J , et al . Decentralized routing and radio resource allocation in wireless ad hoc networks via graph reinforcement learning [J ] . IEEE Transactions on Cognitive Communications and Networking , 2024 , 10 ( 3 ): 1146 - 1159 .

JIA Z Y , HE S J , ZHU Q M , et al . Trusted routing for blockchain-empowered UAV networks via multi-agent deep reinforcement learning [J ] . IEEE Transactions on Communications , 2025 , 73 ( 12 ): 14227 - 14242 .

张雅楠 , 仇洪冰 . 基于深度强化学习的无人机可信地理位置路由协议 [J ] . 电子与信息学报 , 2022 , 44 ( 12 ): 4211 - 4217 .

ZHANG Y N , QIU H B . Trusted geographic routing protocol based on deep reinforcement learning for unmanned aerial vehicle network [J ] . Journal of Electronics & Information Technology , 2022 , 44 ( 12 ): 4211 - 4217 . (in Chinese)

孟超 , 周倩 , 郭林 , 等 . 基于相关性传输模型的无线链路质量估计方法及路由优化算法 [J ] . 电子学报 , 2022 , 50 ( 10 ): 2409 - 2424 .

MENG C , ZHOU Q , GUO L , et al . Estimation method of wireless link quality and routing optimization algorithm based on correlation transmission model [J ] . Acta Electronica Sinica , 2022 , 50 ( 10 ): 2409 - 2424 . (in Chinese)

GANGOPADHYAY S , JAIN V K . A position-based modified OLSR routing protocol for flying ad hoc networks [J ] . IEEE Transactions on Vehicular Technology , 2023 , 72 ( 9 ): 12087 - 12098 .

SONG H , LIU L J , SHANG B D , et al . Enhanced flooding-based routing protocol for swarm UAV networks: Random network coding meets clustering [C ] // IEEE INFOCOM 2021 - IEEE Conference on Computer Communications . Piscataway : IEEE , 2021 : 1 - 10 .

QIU X H , ZHANG S , WANG Z Y , et al . Integrated host- and content-centric routing for efficient and scalable networking of UAV swarm [J ] . IEEE Transactions on Mobile Computing , 2024 , 23 ( 4 ): 2927 - 2942 .

ZHANG L , MA X Z , ZHUANG Z R , et al . Q-learning aided intelligent routing with maximum utility in cognitive UAV swarm for emergency communications [J ] . IEEE Transactions on Vehicular Technology , 2023 , 72 ( 3 ): 3707 - 3723 .

LI J L , XIAO L , QI X C , et al . Reinforcement learning based energy-efficient fast routing for FANETs [J ] . IEEE Transactions on Communications , 2024 , 72 ( 11 ): 7063 - 7076 .

颜志 , 易正伦 , 欧阳博 , 等 . 无人机集群联合拓扑控制的智能路由规划方法 [J ] . 通信学报 , 2024 , 45 ( 2 ): 137 - 149 .

YAN Z , YI Z L , OUYANG B , et al . Intelligent route planning method with jointing topology control of UAV swarm [J ] . Journal on Communications , 2024 , 45 ( 2 ): 137 - 149 . (in Chinese)

惠鏸 , 张晓静 . 无线自组织网络的联合安全路由选择和功率优化算法 [J ] . 电子与信息学报 , 2020 , 42 ( 12 ): 2923 - 2930 .

HUI H , ZHANG X J . Joint secure routing and power optimization algorithm for wireless ad hoc networks [J ] . Journal of Electronics & Information Technology , 2020 , 42 ( 12 ): 2923 - 2930 . (in Chinese)

CAI R J , LI X J , CHONG P H J . An evolutionary self-cooperative trust scheme against routing disruptions in MANETs [J ] . IEEE Transactions on Mobile Computing , 2019 , 18 ( 1 ): 42 - 55 .

LI J L , XIAO L , WANG C X , et al . Learning-based energy-efficient anti-jamming FANET routing with QoS guarantee [J ] . IEEE Transactions on Communications , 2025 , 73 ( 11 ): 11418 - 11431 .

ALTAWEEL A , STOLERU R , GU G F , et al . On detecting route hijacking attack in opportunistic mobile networks [J ] . IEEE Transactions on Dependable and Secure Computing , 2023 , 20 ( 3 ): 2516 - 2532 .

SAKAI K , SUN M T , KU W S , et al . Secure data communications in wireless networks using multi-path avoidance routing [J ] . IEEE Transactions on Wireless Communications , 2019 , 18 ( 10 ): 4753 - 4767 .

杨宏宇 , 韩越 . 基于动态信誉的无线Mesh网络安全路由机制 [J ] . 通信学报 , 2019 , 40 ( 4 ): 195 - 201 .

YANG H Y , HAN Y . Wireless Mesh network secure routing mechanism based on dynamic reputation [J ] . Journal on Communications , 2019 , 40 ( 4 ): 195 - 201 . (in Chinese)

李峰 , 司亚利 , 陈真 , 等 . 基于信任机制的机会网络安全路由决策方法 [J ] . 软件学报 , 2018 , 29 ( 9 ): 2829 - 2843 .

LI F , SI Y L , CHEN Z , et al . Trust-based security routing decision method for opportunistic networks [J ] . Journal of Software , 2018 , 29 ( 9 ): 2829 - 2843 . (in Chinese)

VELUSAMY D , PUGALENDHI G , RAMASAMY K . A cross-layer trust evaluation protocol for secured routing in communication network of smart grid [J ] . IEEE Journal on Selected Areas in Communications , 2020 , 38 ( 1 ): 193 - 204 .

MIRZADEH I , SAYAD HAGHIGHI M , JOLFAEI A . Filtering malicious messages by trust-aware cognitive routing in vehicular ad hoc networks [J ] . IEEE Transactions on Intelligent Transportation Systems , 2023 , 24 ( 1 ): 1134 - 1143 .

HE Y , HAN G J , HOU Y , et al . Environment-tolerant trust opportunity routing based on reinforcement learning for Internet of underwater things [J ] . IEEE Transactions on Mobile Computing , 2025 , 24 ( 7 ): 6348 - 6360 .

HE Y , HAN G J , JIANG J F , et al . CADTR: Context-aware trust routing algorithm based on priority sampling DDPG for UASNs [J ] . IEEE Transactions on Mobile Computing , 2025 , 24 ( 11 ): 11688 - 11702 .

HE Y , FANG J C , YU F R , et al . Large language models (LLMs) inference offloading and resource allocation in cloud-edge computing: An active inference approach [J ] . IEEE Transactions on Mobile Computing , 2024 , 23 ( 12 ): 11253 - 11264 .

CHEN Y F , ZHAO N , DING Z G , et al . Multiple UAVs as relays: Multi-hop single link versus multiple dual-hop links [J ] . IEEE Transactions on Wireless Communications , 2018 , 17 ( 9 ): 6348 - 6359 .

TIMMINS I J , O' YOUNG S . Marine communications channel modeling using the finite-difference time domain method [J ] . IEEE Transactions on Vehicular Technology , 2009 , 58 ( 6 ): 2626 - 2637 .

WANG Z Y , DU J , JIANG C X , et al . UAV-assisted target tracking and computation offloading in USV-based MEC networks [J ] . IEEE Transactions on Mobile Computing , 2024 , 23 ( 12 ): 11389 - 11405 .

SCHWEITZER N , STULMAN A , MARGALIT R D , et al . Contradiction based gray-hole attack minimization for ad-hoc networks [J ] . IEEE Transactions on Mobile Computing , 2017 , 16 ( 8 ): 2174 - 2183 .

段洁 , 闫子豪 , 刘亮 , 等 . 基于多元时变图的天地一体化网络组播路由算法 [J ] . 电子学报 , 2025 , 53 ( 5 ): 1469 - 1481 .

DUAN J , YAN Z H , LIU L , et al . Multicast routing algorithm for space-air-ground integrated networks based on multi-dimensional time-varying graphs [J ] . Acta Electronica Sinica , 2025 , 53 ( 5 ): 1469 - 1481 . (in Chinese)

HUANG G B . Learning capability and storage capacity of two-hidden-layer feedforward networks [J ] . IEEE Transactions on Neural Networks , 2003 , 14 ( 2 ): 274 - 281 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

多智能体强化学习：从基础理论到前沿算法

面向工业场景的边-云协同大语言模型细粒度推理任务卸载

基于动态关系原型的持续关系抽取技术

基于图组合优化的高效社区搜索