Vehicular Ad hoc Networks (VANETs) are one of the most advanced transportation networks that have attracted much attention in recent years. The VANETs are characterized by a large number of traffic flows, which make them a good choice for a wide range of applications. However, due to the unique characteristics of the VANET, routing algorithms present a significant obstacle that must be surmounted. In order to improve the communication quality, the research uses federated learning. The research demonstrates the capacity of the model to learn from its previous errors while also delivering more accurate projections using the federated learning. The findings of the simulation demonstrate that the model with a prediction accuracy of 4 packets/s has the highest accuracy when compared to its contemporaries as well as other predicted models. The results show that the proposed method achieves higher rate of accuracy in transmitting the packets with reduced overhead than the other existing methods.