Masoud Babaeian Dissertation Defense

    Monday, August 5, 2019 - 3:00pm
    Franken Conference Room (Meinel 821)

    Analog Optical Computing Using Nonlinear Optics and Photonics


    Mapping mathematical functions onto the nonlinear photonics operations via all-optical analog approaches suggests several advantages, in terms of energy consumption and speed, over conventional electronic computational platforms.

    In this dissertation, I present simulation and experimental results of all-optical implementations of two important computational problems that currently electronic computational platforms struggle to perform efficiently. 1- Probabilistic graphical models (PGMs) and 2- combinatorial optimization problems (COPs) mapped to a coherent Ising machine (CIM) to find ground state Ising spin configuration of an Ising Hamiltonian. PGMs and COPs are tools that are used to compute probability distributions over large and complex interacting variables. They have applications in social networks, image processing, speech recognition, artificial intelligence, machine learning, and many more areas.

    The optical implementation of PGMs in this work, is based on nonlinear light-matter interaction via a wavelength multiplexing architecture. And the CIM is implemented via a network of injection-locked single frequency, highly doped Yb+3, multicore fiber (MCF) lasers. The overall average accuracy of the implemented CIM to find the ground state Ising spin configuration is ~ 90 % over 120 experimental trials.