Deep Learning based Traffic Aware Network Selection for Enhanced 5G Connectivity

In 5G technology, vehicle networks have vast potential to enhance communication and connectivity. Nevertheless, the challenge of identifying the best network in dynamic and high-traffic areas is still great. Traditional network selection methods are prone to failing in real-time variations, thus causing inefficient use of resources and low user satisfaction. This paper suggests a deep learning-enabled approach to traffic-aware network selection in 5G-enabled vehicular networks with an enhanced security and optimization scheme for Vehicular Ad Hoc Networks (VANETs) to counter new security threats. The Security-Aware Transmission (SESAT) algorithm, with the inclusion of a Malicious Activity Trust (MAT) control system and Elliptic Curve Cryptography (ECC), effectively identifies and thwarts attacks like spoofing and phishing attacks and improves vehicular security. An algorithm based on deep learning using VGG-16 improves network selection by considering mobility patterns and live traffic. The approach ensures cars are connected to the most reliable networks, boosting performance. Accuracy in selection is improved by examining traffic conditions as well as network information. Simulation results show the proposed method vastly outperforms traditional methods through increased accuracy of network selection. It also decreases packet loss significantly, improves general performance, and reduces latency.