Communication Range of Connected Autonomous Vehicles and Its Impact on CO2 Emissions Reduction

This study conceptually examines, using multi-agent simulation-based traffic flow analysis, how the communication range and penetration rate of connected autonomous vehicles (CAVs) influence route choice behaviour, traffic flow distribution, and CO2 emissions. We consider an environment in which CAVs exchange local traffic information and show that traffic congestion mitigation and emission reduction may emerge from the structure of information sharing itself, even without centralised or highly coordinated vehicle control. The results indicate that CO2 emissions respond nonlinearly to the combination of CAV penetration rates and communication ranges, with emission reduction effects becoming evident once penetration exceeds a certain threshold. When communication ranges are limited, relatively high CAV penetration rates are required for such effects to materialise, whereas excessively expanding the communication range is not necessarily desirable, as moderately constrained information sharing can instead promote traffic flow dispersion. By comparison with a reference regime corresponding to user equilibrium under homogeneous information, the analysis suggests that restricting communication ranges introduces spatial heterogeneity in information availability, which can lead to emergent improvements in traffic flow through decentralised route choice. Although individual CAVs act selfishly, the aggregation of local information-based decisions may give rise to swarm-intelligence-like effects at the system level. This study does not aim to predict real-world traffic conditions or emissions, but rather to elucidate the causal mechanisms through which information-sharing structures shape traffic system dynamics under simplified conditions.