With the rapid development of NGSO satellite constellation, multi‐antenna gateway station has become a trend to increase the capacity of communication system. Since the wide use of Ku/Ka bands as well as the dynamic characteristics of NGSO networks, the co‐frequency interference between constellations is inevitable, and the interference limits become important constraints to the system capacity of multi‐antenna gateway stations. In this paper, the allocation of radio resources, including power and beam, under the interference constraints is formulated as an optimization problem. However, traditional optimization methods may not be suitable for NGSO satellite coexistence scenarios due to the fast movements of NGSO satellites. In order to address this problem, the optimization problem is transformed into two subproblems, and a Jensen's Inequality‐based Fast Iteration (JI‐FI) algorithm is designed to solve both of them iteratively. Simulation results show that the interference can be effectively reduced to the value below the threshold and beam and power resources can be allocated simultaneously by adopting the proposed algorithm. Compared with the traditional methods, the proposed scheme can significantly improve the system capacity by more than 25%, and it can greatly reduce the computational complexity.