Wavelength Division Multiplexing in Photonic Reservoir Computing

In this project, I am targeting the realization of brain-inspired information processing systems which make use of multiple wavelengths of light instead of electricity to perform certain computations. The systems can learn from examples and do not need to be explicitly programmed much like how the brain works. They will be integrated in chips in the same way as for example the processor of your phone or computer. Integrated systems are compact, fast, cost-efficient and have a wide variety of possible functions. Using light instead of electrical signals is more power efficient and faster among other advantages. Others have already been looking at such systems but most of them in a non integrated way using bulky components such as optical fibers. There also has not been sufficient research on using multiple wavelengths and the little research that was done focused on using the different wavelengths to perform different tasks at the same time. In my work, apart from integrating the systems on chips, I will take a step further by also looking into using different wavelengths to better perform a single task. The systems will be useful particularly for processing optical data such as in telecom or medical imaging. Examples are processing an internet signal coming from an optical fiber or determining if a passing blood cell is a cancer cell or not.