A Microscopic Traffic Model to Investigate the Effect of Connected Autonomous Vehicles at Bottlenecks and the Impact of Cyberattacks

Bottlenecks reduce both traffic safety and efficiency, resulting in congestion and collisions. The introduction of connected autonomous vehicles (CAVs) has had a significant impact on road networks and can improve traffic efficiency at bottlenecks. This paper proposes a microscopic traffic model to investigate CAV behavior at bottlenecks and examine the effect of cyberattacks. The model is developed using data collected from a roadside sensor node. It is implemented in MATLAB using the Euler scheme to simulate a platoon of vehicles on a circular road of length 1 km. The performance is compared with the intelligent driver (ID) model. The results obtained indicate that the road capacity with the proposed model is 1.4 times higher than with the ID model. Further, the proposed model results in nearly constant speeds with small variations, which is realistic. Conversely, the ID model produces large speed variations that are unrealistic. In addition, the proposed model results in less acceleration and deceleration, which leads to lower vehicle emissions and pollution. The efficiency is better than with the ID model due to CAV communication and coordination, so queues dissipate faster. The traffic flow with the proposed model increases as the density decreases, which is consistent with traffic dynamics. It is also shown that the proposed model can characterize CAV behavior under cyberattacks that cause disruptions in the data. Thus, it can be employed for traffic control and forecasting when bottleneck conditions exist and there is malicious behavior.