Abstract: Aiming at the low spectrum efficiency (SE) in application scenarios of unmanned air vehicles (UAV), an average SE optimization scheme of multi-UAV communication network combined with cognitive radio (CR) was proposed. Firstly,based on cooperative spectrum sensing (CSS), a cognitive UAV network model was established for multi-UAV cooperation in air to ground (A2G) channel, setting the optimization parameters such as number of UAVs, sensing time and cooperative decision threshold. Then,the average SE optimization problem of UAV was investigated, and a joint optimization algorithm was proposed to solve the non-convex optimization problem. Finally, the change of SE in the flight course of UAV was analyzed. The simulation results show that there is an optimal sensing time to maximize the SE, and the number of UAVs and decision threshold will affect the optimal value of SE. In addition, the proposed algorithm has better convergence and can effectively improve the average SE of UAVs in the secondary network.