Based on the research of the human mobilety patterns, a random-walker network model is used to explain the empirical results on the networks of mobile agents, and shows a possible evolution mechanism of these networks. The simulation results show that the degree distribution of this network takes Poisson distribution, and it exhibits the small-world behavior. Moreover, we present a numerical investigation on the displacements of the mobile agents, and find that the distribution of displacements takes power-law. Our numerical results are in good agreement with the empirical results on the human travel from the trajectories of human bank notes.