Abstract:Attack-resilience is essential to maintain continuous service availability in Internet of Vehicles (IoV) where critical tasks are carried out. In this paper, we address the problem of service outage due to attacks on the edge network and propose an attack-resilient mapping of vehicles to edge nodes that host different types of service instances considering resource efficiency and delay. The distribution of service requests (of an attack-affected edge node) to multiple attack-free edge nodes is performed with an optimal vehicle-to-edge (V2E) mapping. The optimal mapping aims to improve the user experience with minimal delay while considering fair usage of edge capacities and balanced load upon a failure over different edge nodes. The proposed mapping solution is used within a deep reinforcement learning (DRL) based framework to effectively deal with the dynamism in service requests and vehicle mobility. We demonstrate the effectiveness of the proposed mapping approach through extensive simulation results using real-world vehicle mobility datasets from three cities.

What problem does this paper attempt to address?

The paper attempts to address the issue of service disruption in the Internet of Vehicles (IoV) caused by attacks on edge networks. Specifically, the authors focus on how to ensure service continuity and availability by optimizing the Vehicle-to-Edge (V2E) mapping during an attack, while maintaining efficient resource utilization and low latency. ### Background and Motivation - **Background**: Edge Networks (EN) in IoV can significantly reduce latency and improve reliability, but they also introduce new security risks such as jamming attacks and denial of service attacks. These attacks can cause edge node failures, affecting vehicle service availability and potentially endangering human lives. - **Motivation**: Existing research mainly focuses on attack defense, with less attention on recovery after edge node failures. Additionally, some studies consider the impact of dynamic and mobile traffic but do not fully address recovery mechanisms post-attack. Therefore, this paper proposes a proactive, attack-resilient V2E mapping method to minimize or avoid service disruption without using backup resources. ### Main Contributions - **Proposed a new mapping model**: Load-Balanced Proactive Secondary V2E Mapping (LB-PSVM), which can remap affected vehicles to non-attacked edge nodes during an attack, while minimizing service latency and maintaining load balance. - **Optimization framework**: The model is based on a Deep Reinforcement Learning (DRL) framework, effectively handling dynamic changes in service requests and vehicle mobility. - **Performance evaluation**: The proposed method's effectiveness is validated through simulations using real vehicle trajectory datasets from three cities. ### Method Overview - **System Model**: The paper proposes a three-layer architecture for an edge-enhanced IoV network, including the data layer, edge layer, and cloud layer. - **Optimization Problems**: - **Primary V2E Mapping**: Under normal conditions, vehicles are mapped to edge nodes providing the required service instances, aiming to minimize latency. - **Secondary V2E Mapping**: During an attack, affected vehicles are remapped to non-attacked edge nodes, aiming to maintain service availability and load balance. - **Service Recovery Placement (SRP)**: After an attack, redeploy affected service instances to ensure network recovery. ### Experimental Results - **Service Latency**: Experimental results show that LB-PSVM effectively maintains low service latency during an attack, with higher resource utilization compared to the Backup Resource (BR) method. - **Load Balancing**: LB-PSVM demonstrates better fairness in load distribution, avoiding overload on individual edge nodes. - **Performance Comparison**: Compared to traditional Secondary V2E Mapping (PSVM) and backup resource methods, LB-PSVM performs better in terms of latency and service quality. ### Conclusion This paper proposes an optimized secondary V2E mapping method that ensures service continuity and availability during edge network attacks, while maintaining efficient resource utilization and low latency. The effectiveness of the method is validated through simulations using real datasets.

Optimizing Vehicle-to-Edge Mapping with Load Balancing for Attack-Resilience in IoV

On Attack-Resilient Service Placement and Availability in Edge-enabled IoV Networks

Adaptive Task Offloading in Vehicular Edge Computing Networks: a Reinforcement Learning Based Scheme

Deep Reinforcement Learning for Load Balancing of Edge Servers in IoV

Vehicle edge server deployment based on reinforcement learning in cloud-edge collaborative environment

Robust Resource Allocation Using Edge Computing for Vehicle to Infrastructure (V2I) Networks

Dynamic Load Balancing for Energy-Delay Tradeoff in a Cloud-RSU-Vehicle Architecture

An Efficient Partial Task Offloading and Resource Allocation Scheme for Vehicular Edge Computing in a Dynamic Environment

An Enhanced AI-Enabled Routing Optimization Algorithm for Internet of Vehicles (IoV)

Correction To: Deep Reinforcement Learning for Load Balancing of Edge Servers in IoV

Task Offloading and Resource Allocation in an RIS-Assisted NOMA-Based Vehicular Edge Computing

DRIVE: Dual rider-remora optimization for vehicular routing

Adaptive Resource Allocation of Vehicles under Dynamic Environment

Computing Resource Allocation Strategy Based on Cloud-Edge Cluster Collaboration in Internet of Vehicles

Deep Reinforcement Learning-Based Task Offloading and Load Balancing for Vehicular Edge Computing

DRLD-SP: A Deep Reinforcement Learning-based Dynamic Service Placement in Edge-Enabled Internet of Vehicles

Intelligent Edge Computing in Internet of Vehicles: A Joint Computation Offloading and Caching Solution

Joint Task Offloading and Resource Optimization in NOMA-based Vehicular Edge Computing: A Game-Theoretic DRL Approach

A Bandwidth-Fair Migration-Enabled Task Offloading for Vehicular Edge Computing: a Deep Reinforcement Learning Approach

DNN Partitioning, Task Offloading, and Resource Allocation in Dynamic Vehicular Networks: A Lyapunov-Guided Diffusion-Based Reinforcement Learning Approach

Internet of Vehicles and Real-Time Optimization Algorithms: Concepts for Vehicle Networking in Smart Cities