Signal Quality Improvement in Optical Communication at Varying Bit Rates using Machine Learning: A Preliminary Study

The long-distance optical communication wavelength-division-multiplexing achieves a 40 Gb/s data rate but needs a dispersion compensator to mitigate dispersion caused by long-distance fiber. The papers suggest employing a multi-layer perceptron neural network model instead to optimize WDM receiver sensitivity and improve signal quality affected by fiber chromatic dispersion.