Applying graph models and Neural Networks on reconfigurable antennas for cognitive radio applications

Abstract

In this paper, graph models and Neural Networks (NNs) are applied to reconfigurable antennas and both techniques are analyzed to be used for cognitive radio applications. The use of NNs to synthesize different antenna configurations and the use of graph models to optimize the performance of such configurations are addressed and investigated. The application of both tools on reconfigurable antennas has proven to be beneficial in applications where software control and fast response are required. Thus the incorporation of such techniques on an FPGA (Field Programmable Gate Array) creates a self adjusting software defined antenna that can be applied to many wireless communication applications such as cognitive radio.