It is of great importance to analyze vehicle trajectory for traffic management and crime investigation using big data mining technology. As a specific mode of the human-data-case, integrating vehicle trajectory information with mobile communication data to study judge cases is important for big data investigation. This paper aims to integrate mobile communication data with traffic road network data, and to provide a theoretical method for analyzing and predicting vehicle trajectory by establishing a reasonable mathematical model and a kind of algorithm. Firstly, the paper presents the problem of vehicle trajectory with mobile communication data, and analyze the topological structure of the road network by graph theory to establish the mathematical model of the road network directed graph. Secondly, the paper applies K-steps Markov chain to establish state transition probability equation of vehicle trajectory model based on historical trajectory data and traffic road network model data. Next, the mobile data is merged into the trajectory of the vehicle. The paper draws the Voronoi Diagram to modify the state transition probability equation of the vehicle trajectory, integrating vehicle trajectory information with mobile communication data.