In this paper, we propose a new approach to route informational messages in VANET. This process contains two steps: the first step is based on the validity of links between vehicles. Using a fuzzy inference system, the available nodes in the network consider the paths with a higher amount of link validity. The second step is inspired by the social behavior of birds through which, by having a population of candidate solutions, a problem is solved. In this step, the available vehicles in the network are divided into some particles. These particles are supposed to find the best path from source to destination in case the congestion is detected and broadcasted by other vehicles. We have considered three parameters as input parameters for PSO, i.e., the output of fuzzy logic, the validity of links, the number of neighbor nodes, and the distance between them. This approach allows PSO to calculate a path for transmitting messages.