Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based Method
Vehicular Edge Computing (VEC) networks have emerged as an efficient paradigm to support a range of computation-intensive applications. However, potential eavesdropping attacks pose significant threats to massive confidential information. Furthermore, the rapid growth of Intelligent Connected Vehicl...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10819378/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841533429568503808 |
|---|---|
| author | Zhen Li Jialong Gong Xiong Xiong Dong Wang |
| author_facet | Zhen Li Jialong Gong Xiong Xiong Dong Wang |
| author_sort | Zhen Li |
| collection | DOAJ |
| description | Vehicular Edge Computing (VEC) networks have emerged as an efficient paradigm to support a range of computation-intensive applications. However, potential eavesdropping attacks pose significant threats to massive confidential information. Furthermore, the rapid growth of Intelligent Connected Vehicles (ICVs) intensifies the scarcity of communication and computational resources, generating an urgent need for improving resource utilization. Since both security concerns and resource limitation influence decisions on offloading strategies and transmission rate, it is necessary to investigate a joint optimization scheme. In this paper, we employ Physical Layer Security (PLS) and design a workflow to address the Joint Secure Offloading and Resource Allocation (JSORA) problem in VEC networks. This workflow models the interaction patterns of multiple ICVs with resource cluster on edge servers, and accurately reflects the occupancy and release of each resource unit. Given the dynamic nature and high complexity of the JSORA problem, we propose a Filtered Deep Reinforcement Learning-based Secure Offloading and Allocation (FDRL-SOA) algorithm to control the offloading and resource allocation within the cluster. Finally, our simulation results demonstrate significant improvements over benchmark methods, with energy consumption reduced by 5.16%, latency decreased by 1.4%, and system cost was minimized by 3.3%. |
| format | Article |
| id | doaj-art-a47b880acd9447128217540323ee3b68 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-a47b880acd9447128217540323ee3b682025-01-16T00:01:26ZengIEEEIEEE Access2169-35362025-01-01134533454610.1109/ACCESS.2024.352463610819378Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based MethodZhen Li0Jialong Gong1https://orcid.org/0009-0007-2299-9061Xiong Xiong2Dong Wang3https://orcid.org/0000-0002-1257-8905China Telecom Research Institute, Beijing, ChinaChina Telecom Research Institute, Beijing, ChinaChina Telecom Research Institute, Beijing, ChinaChina Telecom Research Institute, Beijing, ChinaVehicular Edge Computing (VEC) networks have emerged as an efficient paradigm to support a range of computation-intensive applications. However, potential eavesdropping attacks pose significant threats to massive confidential information. Furthermore, the rapid growth of Intelligent Connected Vehicles (ICVs) intensifies the scarcity of communication and computational resources, generating an urgent need for improving resource utilization. Since both security concerns and resource limitation influence decisions on offloading strategies and transmission rate, it is necessary to investigate a joint optimization scheme. In this paper, we employ Physical Layer Security (PLS) and design a workflow to address the Joint Secure Offloading and Resource Allocation (JSORA) problem in VEC networks. This workflow models the interaction patterns of multiple ICVs with resource cluster on edge servers, and accurately reflects the occupancy and release of each resource unit. Given the dynamic nature and high complexity of the JSORA problem, we propose a Filtered Deep Reinforcement Learning-based Secure Offloading and Allocation (FDRL-SOA) algorithm to control the offloading and resource allocation within the cluster. Finally, our simulation results demonstrate significant improvements over benchmark methods, with energy consumption reduced by 5.16%, latency decreased by 1.4%, and system cost was minimized by 3.3%.https://ieeexplore.ieee.org/document/10819378/Vehicular edge computingsecuritydeep reinforcement learning |
| spellingShingle | Zhen Li Jialong Gong Xiong Xiong Dong Wang Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based Method IEEE Access Vehicular edge computing security deep reinforcement learning |
| title | Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based Method |
| title_full | Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based Method |
| title_fullStr | Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based Method |
| title_full_unstemmed | Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based Method |
| title_short | Multi-Slot Secure Offloading and Resource Management in VEC Networks: A Deep Reinforcement Learning-Based Method |
| title_sort | multi slot secure offloading and resource management in vec networks a deep reinforcement learning based method |
| topic | Vehicular edge computing security deep reinforcement learning |
| url | https://ieeexplore.ieee.org/document/10819378/ |
| work_keys_str_mv | AT zhenli multislotsecureoffloadingandresourcemanagementinvecnetworksadeepreinforcementlearningbasedmethod AT jialonggong multislotsecureoffloadingandresourcemanagementinvecnetworksadeepreinforcementlearningbasedmethod AT xiongxiong multislotsecureoffloadingandresourcemanagementinvecnetworksadeepreinforcementlearningbasedmethod AT dongwang multislotsecureoffloadingandresourcemanagementinvecnetworksadeepreinforcementlearningbasedmethod |