Abstract:The current usage of federated learning in applications relies on the existence of servers. To address the inability to conduct federated learning for IoV (Internet of Vehicles) applications in serverless areas, a P2P (peer-to-peer) architecture for federated learning is proposed in this paper. Following node segmentation based on limited subgraph diameters, an edge aggregation mode is employed to propagate models inwardly, and a mode for propagating the model inward to the C-node (center node) while aggregating is proposed. Simultaneously, a personalized differential privacy scheme was designed under this architecture. Through experimentation and verification, the approach proposed in this paper demonstrates the combination of both security and usability.

What problem does this paper attempt to address?

The paper primarily proposes a new solution for federated learning in the Internet of Vehicles (IoV) environment. Specifically, the paper attempts to address the following key issues: 1. **Federated Learning Application in the Absence of a Server**: Existing federated learning methods typically rely on the presence of a central server to coordinate the training process among various clients. However, in the IoV environment, especially in areas where servers are absent or difficult to deploy, traditional federated learning methods are challenging to implement. Therefore, the paper proposes a peer-to-peer (P2P) architecture-based federated learning method to overcome this limitation. 2. **Node Distance and Communication Efficiency**: In the IoV scenario, the distance between vehicle nodes varies, which may lead to low communication efficiency or even ineffective communication. To address this issue, the paper proposes a distance-based node grouping method, which divides vehicle nodes that are geographically close into the same subgraph, thereby forming a local P2P network. 3. **Balancing Privacy Protection and Model Quality**: Although federated learning itself has a certain degree of privacy protection, there is still a risk of data leakage, especially when malicious nodes can infer sensitive information through model parameters. To enhance privacy protection, the paper designs a personalized differential privacy scheme that dynamically adjusts the amount of noise added to the model parameters based on the geographical distance between nodes. When the distance between nodes is short, more noise is added to enhance privacy protection; conversely, less noise is added to maintain the quality of the model. In summary, the main contributions of the paper include: - Proposing a distance-segmented node grouping algorithm (F-Prim algorithm) to control the subgraph diameter, thereby constructing a P2P federated learning architecture suitable for the IoV environment. - Designing a hierarchical propagation path that allows model parameters to aggregate from edge to central nodes within the subgraph. - Introducing a distance-adjusted personalized differential privacy scheme aimed at balancing privacy protection and model quality.

P2P Federated Learning Based on Node Segmentation with Privacy Protection for IoV

Decentralized Iot Data Sharing: A Blockchain-Based Federated Learning Approach with Joint Optimizations for Efficiency and Privacy

Privacy protection in intelligent vehicle networking: A novel federated learning algorithm based on information fusion

Blockchain-Based Federated Learning in Mobile Edge Networks with Application in Internet of Vehicles

Decentralized Proactive Model Offloading and Resource Allocation for Split and Federated Learning

Risk-Aware Reinforcement Learning Based Federated Learning Framework for Io V

Privacy-Preserved Federated Learning for Autonomous Driving

Secure Decentralized Aggregation to Prevent Membership Privacy Leakage in Edge-based Federated Learning

Federated Split Learning with Data and Label Privacy Preservation in Vehicular Networks

Edge-assisted U-Shaped Split Federated Learning with Privacy-preserving for Internet of Things

A Study of Federated Learning with Internet of Things for Data Privacy and Security using Privacy Preserving Techniques

Two-Phase Multi-Party Computation Enabled Privacy-Preserving Federated Learning

Collaborative IoT learning with secure peer-to-peer federated approach

LSFL: A Lightweight and Secure Federated Learning Scheme for Edge Computing

Decentralized P2P Federated Learning for Privacy-Preserving and Resilient Mobile Robotic Systems

Joint Top-K Sparsification and Shuffle Model for Communication-Privacy-Accuracy Tradeoffs in Federated Learning-Based IoV

Peer-to-peer privacy-preserving vertical federated learning without trusted third-party coordinator

Federated Learning-based Privacy-aware Location Prediction Model for Internet of Vehicular Things

PPT: A Privacy-Preserving Global Model Training Protocol for Federated Learning in P2P Networks

Keep Your Data Locally: Federated-Learning-Based Data Privacy Preservation in Edge Computing

PSDF: Privacy-aware IoV Service Deployment with Federated Learning in Cloud-Edge Computing