Vehicle networks have become a prominent research topic because of their distinctive characteristics and uses, such as consistency, effective traffic control, road security, and infotainment. The network entities must make judgments on how to exploit network operation under ambiguous conditions. A challenging environment is produced by the expanding usage of wireless technology in a highly mobile environment. To improve communication dependability in this environment, intelligent technologies must be used to solve the routing issue and construct a more durable communication system. An excellent solution to this problem is reinforcement learning (RL). You can achieve your goal by using reinforcement learning (RL), which can effectively handle challenges with decision-making. Yet, the state and action spaces in large-scale wireless networks are huge and complex. As a result, it's possible that RL won't be able to decide on the best course of action in time. Deep reinforcement learning (DRL), a hybrid of RL and DL, was created as a solution to this issue. We begin by introducing vehicular networks and giving a brief overview of the concepts of RL and DRL in this analysis. Then, to address fresh issues in 6G vehicle networks, we discuss RL and, in particular, DRL approaches. Finally, we list a few concerns that still need to be studied further.