Abstract: Aiming at the multi-objective optimization problems such as coverage rate, communication quality and flight energy consumption existing in the deployment process of 3D space mobile network base stations, we propose an improved multi-objective whale optimization algorithm. Firstly, a spatial channel model is constructed by integrating the 3D spatial deployment characteristic to form a multi-objective optimization problem for the deployment task. Secondly, the K-medoids algorithm is utilized to improve the initial position of unmanned aerial vehicle (UAV) network base station, accelerating the iteration speed, avoiding invalid iterations and enhancing the quality of solution. Thirdly, the improved sine and cosine perturbation factors are integrated into the spiral search to enhance the population diversity and escape the local optimum. Finally, the golden section coefficient is introduced to adjust the search direction, search for the optimal individual more accurately, and improve the solution ability and convergence speed. The compared simulation results show the effectiveness of the algorithm proposed under three distributions in multiple scenarios, and it has obvious advantages in coverage rate and flight energy consumption.