Wavelength division multiplexing reservoir computer using quantum dot lasers with delayed optical feedback

Time-delay reservoir computer (RC) based on semiconductor lasers provides a simple hardware implementation of the recurrent neural network. However, the data processing speed is limited by the length of the feedback loop. This work demonstrates a parallel RC scheme based on the wavelength division multiplexing (WDM) technique. This scheme is implemented on a Fabry-Perot quantum dot laser with multimode emission. It is shown that the four-channel WDM RC exhibits a better performance over the single-channel one, with the same number of virtual neurons. Meanwhile, the RC is accelerated by four times, owing to the shorter delay time. In addition, we show that the cross-gain saturation effect between the multimodes plays a crucial role on the RC performance.