Deep learning-enabled quality of service assurance for real-time multimedia in dynamic IoV networks

Purpose With the rapid development of the Internet of Vehicles (IoV), vehicles will become smarter and safer with high-quality multimedia communications and real-time connectivity. This study aims to address the challenges of dynamic resource allocation, high mobility and heterogeneous network conditions in IoV environments by using artificial intelligence (AI) to optimize quality of service (QoS). Design/methodology/approach In this framework, advanced AI models are used to select and manage access points to enhance multimedia communication. For real-time multimedia applications, this study introduces a Double Deep Q-Learning (DDQN) model that minimizes network delays, increases energy efficiency and enables high throughput and low latency. Findings Based on extensive evaluations, the proposed framework outperforms both traditional and existing AI-based methods in QoS provisioning. Originality/value The use of advanced AI models, particularly the DDQN model, in this study, represents an innovative approach to improving QoS in IoV environments. This makes the proposed framework an excellent solution for the next generation of Internet of Things networks.