Federated Learning and Blockchain-Enabled Framework for Traffic Rerouting and Task Offloading in the Internet of Vehicles (IoV)
Ganesh Gopal Devarajan, Thangam S, Mohammed J. F. Alenazi, U Kumaran, Gopalakrishnan Chandran, Ali Kashif Bashir
Abstract
The Internet of Vehicles (IoV) presents significant opportunities for enhancing traffic management and vehicle coordination, but it also faces challenges related to traffic congestion, data privacy, and efficient computational resource allocation. Traffic congestion remains a critical problem, impacting travel time, fuel consumption, and emissions. Additionally, task offloading in the edge-cloud environment demands efficient strategies to balance latency, resource usage, and computational load. Our proposed system, Joint Federated Learning and Blockchain-Enabled Traffic Rerouting with Efficient Task Offloading of Consumer IoV in the Edge-Cloud Environment, addresses these issues by integrating federated learning and blockchain technologies. Federated learning allows vehicles to collaboratively train a global model without sharing raw data, preserving privacy and reducing bandwidth usage. Blockchain ensures the security and integrity of the model updates, fostering trust among participants. Efficient task offloading strategies optimize the use of edge and cloud resources, minimizing latency and energy consumption. Our approach is validated using a comprehensive dataset, and the results demonstrate significant improvements in traffic prediction accuracy, security, and overall system performance, highlighting the effectiveness of the integrated solution in addressing the challenges of Consumer Internet of Vehicles (CIoV).