Offshore wind farms are increasingly becoming the focus of clean sources market because of the huge energy potential and fast-maturing technology. The existing researches normally optimize the wind turbine layout and two-dimensional cable routing independently. This work focuses on the synchronization optimization of site selection of the offshore wind farm, three-dimensional wind turbine layout and three-dimensional cable network routing based on meta-heuristic algorithms and geographic information systems. Several practical issues, i.e., restricted areas, power generation, cable network and energy loss, are taken into consideration. A two-layer model is proposed. The outer layer model is for the site selection and the wind turbine layout optimization. The inner layer model is for the obstacle-avoiding cable routing optimization. In this stage, the seabed terrain is considered for the first time. The proposed integrated model is complex and non-convex. Thus, a hybrid method including an improved ant colony optimization combined with genetic algorithm, dual-simplex method and Kruskal algorithm is proposed to search the solution more efficiently. The initialization stage of the hybrid method is improved from random assignment to directional assignment. The directional solution is obtained by the widely used genetic algorithm. A case study based on a real offshore wind farm is established to prove the effectiveness of the proposed methodology. The results show an over one million dollars increase in annual benefit compared with conventional methods.