The Effects of Vehicle-to-Infrastructure Communication Reliability on Performance of Signalized Intersection Traffic Control

Abstract

A vehicle-to-infrastructure communication can inform an intersection controller about the location and speed of connected vehicles. Recently, the design of adaptive intersection control algorithms that take advantage of this information evoked a lot of research, typically assuming a perfect communication. In this study, we explore possible effects of a temporal decrease in the reliability of the communication channel, on a throughput of a signalized intersection. We model road traffic and DSRC-VANET communication by integrating the well-established traffic and communication simulation tools (Vissim and OMNeT++). Comparisons of the perfect communication conditions with challenging, but realistic conditions involving communication distortions show that the adaptive intersection control may suffer from significantly increased average delays of vehicles. The level of delays is largely independent of whether the communication distortions affect all or only a single intersection approach. For some signal groups, the average vehicle delays are significantly increased, while for the others they are decreased, leading to the disbalance and unfairness. We show that the data received in the previous time intervals and simple assumptions about the vehicle movements can be used to estimate the lost data. The compensation of the communication distortions affecting a single intersection approach decreases the average vehicle delays. When the communication distortions impacting all the intersection approaches are compensated for, the vehicle delays are even set back to the levels comparable with the perfect communication conditions. Overall, the results demonstrate that the impact of the communication distortions should be considered in the design of the adaptive intersection control algorithms.