Summary In vehicular ad hoc network (VANET) broadcasting is considered as the critical area of research. The vehicles are connecting in an ad hoc manner to create a network of a wider range. In an intelligent transport system (ITS), with vehicle to vehicle (V2V) communication in the VANET is used to support the network backbone and which the accident prevention technique has been deployed for ensuring the road safety. The major reason for the road accident is the abnormal driving behavior of human drivers. The movement pattern of the vehicle is the main factor of the network topology so that the impact of human driving pattern influences the performance and behavior of the network. Driving behavior can be classified as reckless, normal, drunken, and fatigue driving. During driving, the behavior of human drivers was predicted in this paper, thereby providing the performance analysis on the mobility of the network. For the improvement of a driver behavior, the hybridized mega‐trend diffusion (MTD) with optimized deep belief network–sunflower optimization (DBN‐SFO) is used. Here, the network performance of the adaptive deer hunting Optimized Link State Routing Protocol (ADHOLSR) is analyzed in NS2 simulation platform.