Controlling reconfigurable antennas via Neural Network embedded into an FPGA

In this paper, a software controlled (SC) field programmable gate array (FPGA) controller is designed and shown that it can be used to control switches on a reconfigurable antenna based on a Neural Network (NN) intelligent algorithm. The SC-FPGA controller is designed by training a NN model using customizable blocks designed in the MatLab Simulink and Xilinx System Generator. Network parameters are mapped into a hardware structure that improves the performance and the efficiency. Examples of how this controller can be embedded into an antenna system will be shown and discussed.

[1]  T. H. O'Donnell,et al.  Direction finding in phased arrays with a neural network beamformer , 1995 .

[2]  J. Costantine,et al.  A frequency reconfigurable rotatable microstrip antenna design , 2010, 2010 IEEE Antennas and Propagation Society International Symposium.

[3]  Qi-Jun Zhang,et al.  Neural Networks for RF and Microwave Design , 2000 .

[4]  Ivan Grech,et al.  Artificial Neural Network Optimization for FPGA , 2006, 2006 13th IEEE International Conference on Electronics, Circuits and Systems.

[5]  James Lyke.  A cellular automata FPGA architecture that can be trained with neural networks , 2002, Proceedings, IEEE Aerospace Conference.

[6]  Kuan-Kin Chan,et al.  A neural network approach to MVDR beamforming problem , 1992 .

[7]  Michael Georgiopoulos,et al.  Applications of Neural Networks in Electromagnetics , 2001 .