The growing demand of the radio spectrum is an important part in the multi-agent intelligence management system of the vehicles. Cognitive radio is used for reducing the restricted access to the wavelength of the spectrum and utilizing the radio spectrum is dynamic allocation method. With the advent in the cognitive radio arrangement, the CR in vehicular ad hoc networks allow the operator to sense and hop from one to another system network in the desired frequency of the spectrum based on the environment of the cognitive radio. In existing method implemented a cluster formation mechanism used for data transmission one to another vehicle nodes. In this mechanism used CR-VANET network is divided into subgroups or clusters and achieve accuracy rate among vehicles. In this work, has implemented a cluster formation mechanism with Bacterial foraging optimization algorithm method in Cognitive radio in VANETs. In planned technique a self-motivated system network is established on the basis of clusters using BFOA network goals to achieve better throughput in data transmission one node to another node with RSU (Road Side Units). In the experimental result improves accuracy of the data transmission over the network. In proposed research, vehicles and road side units are deployed in the network. When there is loss of the data packets during the transmission in the network, then optimized clustering phase has implemented. In addition, the selections of the cluster heads are maintained the path and optimization (BFOA) phase implement to recover the path losses and improve the network performance such as overhead, energy consumption, E2E delay and Network Throughput and compared with existing method (Cluster-Formation). Simulation tool used in this proposed work is MATLAB 2016a.