Comparison between the Interconnection and Self-connection Optical Neural Networks on the Application of AND and OR Logic Gates

Chu-En Lin^1*, Ching-Pao Sun¹, Chii-Chang Chen¹

¹光電科學與工程學系, 國立中央大學, 桃園市, Taiwan

* Presenter:Chu-En Lin, email:celin2021@cc.ncu.edu.tw

In this study, we use two configurations to research the performance for AND and OR logic gate. Through our results, the AND and OR Logic Gates present the best performance when the input signals modulated by intensity modulator with interconnection configuration. Under these conditions, the best bit error ratio (BER) and Normalized root mean square error (NRMSE) for AND gate are 0.1548 and 0.53, respectively. For OR gate, the BER and NRMSE are 0.0654 and 0.51, respectively.

Keywords: Optical neuron networks, Integrated optical device, Reservoir computing