Abstract: A fast computation method of mutual coupling inverse matrix in uniform linear antenna array based on RBF neural network is proposed to solve the problems of computational complexity and difficult to implement in real-time in traditional method. Different training methods can be applied according to different calculation methods for mutual coupling inverse matrix. The mutual coupling inverse matrix can be estimated in real-time owing to the RBFNNs superior ability of parallel processing, and the computation problem can be transformed to nonlinear function approximation. The specific steps are presented. And the simulation results prove its superiority to other methods.