Deep Learning-Based Prediction of Multifunctional Photonic Crystal Ring Resonator With Ultra High-Quality Factor

A novel deep learning-based reconfigurable and multifunctional Photonic Crystal Ring Resonator (PCRR) is designed with narrow bandwidth, low insertion loss and ultracompact size for lightwave communication and optical computing applications. The designed coupled nanoring resonator is used to realize four different functions of optical switch, narrow bandpass filter, encoder and XOR gate. The periodic structure of photonic bandgap frequency range is calculated by the Plane Wave Expansion (PWE) technique. The multifunctional nanoscale structure performance parameters of extinction ratio, quality factor and insertion loss are numerically analyzed by Finite-Difference-Time-Domain (FDTD) method. The deep learning algorithm of Long Short Term Memory- Neural Network (LSTM-NN) is used to predict the design parameters with low mean square error and less computation time of 50 seconds. The nanoring resonators is designed with high quality factor of 2566.83, high extinction ratio of 34.87 dB and ultracompact size of 179.20 μm2. Hence, this multifunctional platform is highly appropriate for photonic integrated circuits and optical computing system.