AI-based efficient wireless technologies and infrastructure-based networks with VANET for smart transportation high performance

Abstract

VANETs (Vehicular Ad-hoc NETworks) were deemed the most suitable communication network for supporting the dissemination of alert messages due to their low dissemination delays as well as extensive vehicle coverage in vicinity of an emergency. With the introduction of cooperative ITS services, it is envisaged that emerging vehicular networks will progressively rely on Vehicle to Infrastructure (V2I) communication lines, which are expected to be nominally accessible with certain temporary as well as time-limited connectivity losses. This study proposes a novel method for VANET-based efficient vehicle clustering and routing based on network infrastructure for high-performance smart transportation. The vehicle clustering is using infrastructure-based fuzzy K-means convolutional neural networks. Then the energy-efficient cluster-based multi-hop distributed routing has been carried out. The experimental analysis in terms of latency, network lifetime, throughput, QoS, energy efficiency, and packet delivery ratio. In addition, empirical equations that can be used to predict speed recommendations for drivers are derived from the result.