Currently, the heterogeneous vehicular network becomes a very popular topic in the connected vehicle field since it can take advantage of multiple wireless networks to provide the best connection service anytime and anywhere. There were many studies focused on optimizing the parameters such as propagation delay, packet delivery rate, etc., to improve the network performance or meet the preferences of various vehicular network applications, but few studies have been aimed at building models to evaluate optimizations so far. Some previous works have also regarded the network performance as constant, which results in instability when they are used in real-world scenarios with varying network performance. To solve the above problems, we propose a novel model named Capacitated Static Heterogeneous Vehicular Network Assignment (CSHVNA) model. We first propose a heterogeneous vehicular network structure for distributed network selection. Then, System Optimum (SO) and User Equilibrium (UE) are regarded as ideal conditions of system-centric network selection methods and user-centric network selection methods respectively. The uniqueness of the solution to the model is also verified. The established model refers to the static traffic assignment model. Finally, three network selection methods, namely, a greedy algorithm-based method, an evolutionary game-based method, and a two-layer game-based method are presented to validate the proposed model. The simulation results show that the greedy algorithm-based method will cause a serious ping-pong effect in the capacitated network model. The evolutionary game-based method can achieve UE while the two-layer game which uses prior knowledge can approximate SO. The proposed model is validated through the numerical simulation experiment.

CICTP