Driven Communication Congestion in Connected Vehicle Networks: A Complexity Metrics Approach

Research on connected vehicle network congestion has predominantly focused on congestion control algorithms and connectivity optimization, while systematic investigation into the interplay between network topology evolution and communication performance under complex traffic conditions remains limited. Existing studies lack in-depth analysis based on complex network metrics, hindering the development of robust frameworks for predicting and managing communication congestion. This paper addresses this gap by employing a multi-methodological approach guided by complex network theory. Using the Simulation of Urban Mobility (SUMO) tool, we construct two representative traffic scenarios encompassing six distinct traffic conditions. Through graph-theoretic analysis, we quantify four key network complexity metrics: node degree distribution, degree-degree correlation, average path length, and betweenness centrality. We then apply statistical modeling to dissect the mechanistic relationships between traffic density, vehicle communication range, and the structural evolution of connected vehicle networks. Building on these empirical findings, we establish quantitative links between these topological features and patterns of communication congestion, offering new theoretical foundations for congestion prediction and proactive network management in connected vehicle systems.