Fig. 1. Schematic diagram of micro-ring resonator structure

Fig. 2. Structure diagram of Kerr microring based photonic reservoir computing system

Fig. 3. Characteristic analysis of pure Kerr microring resonators. (a) Stability results obtained through steady-state solutions and perturbation methods; (b) dynamic response under 100 MHz modulation frequency

Fig. 4. Simulation results of photonic reservoir computing based on Kerr microring resonator in speech task. (a) The first sample preprocessed via Lyon’s cochlear model and converted into signals with 40 distinct frequency channels and 40 time steps; (b) reservoir responses of channel 19 under different input rates; (c) responses of microring reservoir across different channels at an input sampling rate of 200 Gbit/s; (d) confusion matrix of test dataset under a 200 Gbit/s input signal rate with 4 virtual nodes per channel

Fig. 5. Dependence of classification accuracy on the number of virtual nodes with 200 Gbit/s input signal

Fig. 6. Performance of the reservoir under Nv=4 and different input rate