Modeling of Nano-Contact Vortex Oscillator Using Recurrent Neural Network
Résumé
NanoContact Vortex Oscillator (NCVO) is a highly nonlinear electronic device that can depict chaotic or nonchaotic behaviors according to the applied DC bias current or the magnetic field. The potential use of such device in computing systems requires the build of a realistic model capable of describing its dynamics. In this paper, we demonstrate an efficient approach for the modeling of the NCVO device based on a recurrent neural network. The proposed model compared to micromagnetic simulations at different current values, shows promising results.