Multi-Wavelength Parallel Training and Quantization-Aware Tuning for WDM-Based Optical Convolutional Neural Networks Considering Wavelength-Relative Deviations

Wavelength Division Multiplexing (WDM)-based Mach-Zehnder In-terferometer Optical Convolutional Neural Networks (MZI-OCNNs) have emerged as a promising platform to accelerate convolutions that cost most computing sources in neural networks. However, the wavelength-relative imperfect split ratios and actual phase shifts in MZIs and quantization errors from the electronic configuration module will degrade the inference accuracy of WDM-based MZI-OCNN s and thus render them unusable in practice. In this paper, we propose a framework that models the split ratios and phase shifts under different wavelengths, incorporates them into OCNN training, and introduces quantization-aware tuning to maintain inference accuracy and reduce electronic module complexity. Consequently, the framework can improve the inference accuracy by 49%, 76%, and 76%, respectively, for LeNet5, VGG7, and VGG8 implemented with multi-wavelength parallel computing. And instead of using Float 32/64 quantization resolutions, only 5,6, and 4 bits are needed and fewer quantization levels are utilized for configuration signals.