A new field called Vehicular Adhoc Networks (VANETs) uses wireless local area networks (WLANs) with an ad-hoc topology. Routing complexity and high control overhead are two frequent challenges faced by vehicular ad hoc networks (VANETs). However, most of these initiatives failed to provide a comprehensive solution to the problems associated with routing and control overhead minimization. In order to lower the augmented control overhead, the current work presents an Improved Deep Reinforcement Learning (IDRL) method for routing. As 5G cells arrive, emerging automotive networks could make driving safer, more environmentally friendly, and more efficient. They should also pave the road for autonomous driving. High sequence factors in vehicle settings give rise to a variety of new irritants, which is why remote design approaches are being reconsidered. Future smart cars, which constitute the foundation of high-performance multipurpose networks, are steadily receiving increasingly sophisticated sensors and are still generating vast amounts of data.