BDEdge: Blockchain and Deep-Learning for Secure Edge-Envisioned Green CAVs
Prabhat Kumar, Randhir Kumar, Govind P. Gupta, Rakesh Tripathi
Abstract
Green Connected and Autonomous Vehicles (CAVs) are the future of next-generation Intelligent Transportation Systems (ITS) that will help humans to improve road safety and reduce pollution, energy consumption, and travel delays. To increase the performance of green CAV, the data generated by Autonomous Vehicles (AVs) and associated infrastructure needs to be processed in real-time. Mobile Edge Computing (MEC) is a promising paradigm that can be integrated with green CAV to save energy and to improve the network performance in terms of low latency for data processing. However, MEC servers and other communication entities in green CAV environment cannot be fully trusted and may bring vulnerabilities related to data privacy and security. Motivated by the above challenges, we design a blockchain and Deep-Learning (DL)-enabled secure data processing framework for an edge-envisioned green CAV environment (hereafter referred to as BDEdge). In blockchain-based scheme, all communication entities are registered, verified and thereafter validated using smart contract-based Practical Byzantine Fault Tolerance (PBFT) consensus algorithm. The authenticated data is forwarded to DL scheme. In DL-based scheme, an Intrusion Detection System (IDS) based on a hybrid model of Sparse Auto-Encoder-enabled Attention Bidirectional Gated Recurrent Unit (SAE-ABIGRU) is designed by analyzing the link load behaviors of the MEC-enabled Road Side Unit (RSU) server. The security analysis and comparative simulation findings shows that the proposed BDEdge framework can significantly reduce false alarm rate and increase accuracy close to 99%.