From Network Topology to Traffic: An investigation of the Relationship Between Network Topological Features and Traffic Data

Abstract

Traffic dynamics is a complex phenomenon. Visualizing and uncovering its hidden complexities using simple topological features with clear physical interpretations allows us to improve traffic predictions and offer valuable insights to decision-makers. This study delves into the relationship between these topological features and actual traffic data, with an emphasis on the highway network in the Netherlands. Using both traffic data and complex network analysis techniques, we explore how the betweenness centrality (BC) corresponds to patterns in traffic flow and speed. We apply Pearson correlation analysis to quantify these relationships, especially during peak traffic hours. Interestingly, while the results show that the correlation between BC and traffic flow and speed is not strong during the day, a more intricate relationship emerges during peak times. We showed that BC demonstrates a notable negative correlation with traffic speed, a finding that is statistically significant (p-value ≤ 0.05). These insights pave the way for a deeper understanding of how network topology affects traffic behavior.