Internet of Vehicles (IoV) is regarded as an emerging paradigm for connected vehicles to exchange their information with other vehicles using vehicle-to-vehicle (V2V) communications by forming a vehicular ad hoc networks (VANETs), with roadside units using vehicle-to-roadside (V2R) communications. IoV offers several benefits such as road safety, traffic efficiency, and infotainment by forwarding up-to-date traffic information about upcoming traffic. For instance, IoV is regarded as a technology that could help reduce the number of deaths caused by road accidents, and reduce fuel costs and travel time on the road. Vehicles could rapidly learn about the road condition and promptly respond and notify drivers for making informed decisions. However, malicious users in IoV may mislead the whole communications and create chaos on the road. Data falsification attack is one of the main security issues in IoV where vehicles rely on information received from other peers/vehicles. In this paper, we present data falsification attack detection using hashes for enhancing network security and performance by adapting contention window size to forward accurate information to the neighboring vehicles in a timely manner (to improve throughput while reducing end-to-end delay). We also present clustering approach to reduce travel time in case of traffic congestion. Performance of the proposed approach is evaluated using numerical results obtained from simulations. We found that the proposed adaptive approach prevents IoV from data falsification attacks and provides higher throughput with lower delay.