Analysis of traffic congestion by network community structure

A novel model is proposed to characterize the dynamics of city traffic states, especially traffic congestion, based on the concept of network community structure. Concretely, using the Floating Car Data (FCD) information of vehicles, we construct a new traffic network model with floating cars as network nodes. Taking the traffic of Beijing city as an instance, comparing with free traffic cases, our simulations show that the traffic network generally has more prominent community structures with larger values of network modularity for congested traffic situations. Our results show that the network community structure is a principled tool for investigating city traffic congestion problems.