Semiconductor lasers as reservoir substrates
Van der Sande, Guy; Soriano, Miguel C.
Photonic Reservoir Computing: Optical Recurrent Neural Networks, De Gruyter, 185-203 (2019)
Semiconductor lasers are the most common type of lasers produced, with a wide range of applications that include optical storage systems, communication systems (ranging from short-distance data communication systems to long-haul fiber-optic networks), as pump sources, for material processing, and in many other applications. A recent application, and the topic of this chapter, is the use of semiconductor lasers as substrates for photonic reservoir computers. Here, we present an overview of recent publications and discuss the main properties of photonic reservoir computers based on semiconductor lasers subject to optical feedback. The optical feedback serves as a way to create the recurrence required for reservoir computing in the context of delay-based reservoir computers.