AI-Driven Secure Emergency Message Dissemination in 5G-Enabled VANETs Using LSTM-Based Intrusion Detection and CP-ABE Encryption

Abstract

The number of vehicles has expanded rapidly due to advances in automobile technology and global population growth, resulting in an increase in the frequency of traffic accidents. Wireless Vehicle-to-Vehicle (V2V) connections are used by event-driven safety applications to alert drivers to potentially dangerous situations. For emergency vehicles to respond to urgent emergency services, there must be uninterrupted traffic on the roads. Even a small delay in an emergency journey time can be costly and potentially result in lost lives. To overcome this limitation in this research, we designed an AI-driven emergency message dissemination framework for “Vehicular Ad hoc Networks (VANETs)” with 5G, for the efficient, secure, and privacy-preserving communication of messages at critical times. Using time-series data analysis, an AI-based Intrusion Detection System (IDS) that uses Long Short-Term Memory (LSTM) networks classifies emergency messages through recognizing abnormalities and false warnings. To ensure secure emergency message dissemination in the VANET environment, this research proposes a Multi-Authority Ciphertext-Policy Attribute-Based Encryption (CP-ABE). Priority-based scheduling leverages are the Edge-DENM Prioritization Algorithm, which categorizes emergency messages based on urgency levels to prevent delays in high-risk scenarios. The proposed approach ensures secure, scalable, and privacy-preserving emergency communication, improving overall VANET performance by dynamically adapting to traffic conditions and vehicle mobility.