Antenna array synthesis based on pre-process coding neural network

Antenna array synthesis based on the back propagation neural network is proposed in this Letter. The synthesis process is divided into three steps: pre-processing, training, and post-processing. In the pre-processing, the interference directions and desired signal direction are coded so as to synthesise an arbitrary number of interferences. The back propagation neural network is unified to emulate the convex array synthesis model. To verify the performance of the network, the nine-element and 11-element linear array synthesis based on the neural network is realised and tested. Synthesis results show that the proposed method gives accurate approximation and have the good generalisation ability.