Thomas - Adaptive Optics Using Machine Learning

Abstract

Lasers are becoming increasingly important to the US Navy due to possibility of a near peer conflict where traditional communications across the electromagnetic spectrum can be easily be intercepted or jammed. Lasers can be used for communication in free space optical communications systems. Free space optical communication systems allow for wireless communications that can neither be intercepted nor jammed by the enemy. The maritime environment degrades laser light propagation through optical turbulence, and this can make free space optical communication systems less effective. Our research will focus on developing an adaptive optics system that is able to correct for the effects of optical turbulence on laser propagation. Previous research has been able to recreate simulated optical turbulence that allowed for optimization of laser light intensity delivered on target. Our research will build on previous work experimentally correcting for the effects of optical turbulence on laser propagation. We will build on previous work by predicting and correcting for optical turbulence in air and underwater. Zernike polynomials will be projected onto a spatial light modulator (SLM) to recreate the effects of optical turbulence. Machine learning will then be used to determine the specific Zernike polynomial that corresponds to a specific optical turbulence strength. Once a Zernike polynomial is matched to a specific optical turbulence strength, the complex conjugate of that Zernike polynomial will then be projected onto another SLM in order to correct for the effects of optical turbulence.

Motivation

• Naval uses for free space optical communication

• Adaptive optics will help optimize free space optical communications