Pioneering Work in Optical Computing: Deciphering the “Undecipherable”

A group of researchers has identified a technique to unravel information that gets mixed up when light travels through a dispersing medium, like ground glass. This finding can be beneficial in the realms of optical computing and machine learning. The team showcased that the input-output optical response of a nonlinear scattering medium could be depicted as a third-order tensor, introducing a fresh perspective on optical encryption and logic gates.

Interpreting the Response of Nonlinear Scattering Media: A Giant Leap Towards Highly Scalable Physical Operators

Has it ever been thought possible to view through a scattering medium such as ground glass? This idea has been generally seen as unachievable. As light transits through an opaque substance, the information it carries gets confused, akin to it being subjected to intricate encryption.

Professor Choi Wonshik’s team from the IBS Center for Molecular Spectroscopy and Dynamics (IBS CMSD) recently made an astounding scientific development, revealing a technique to exploit this phenomenon in optical computing and machine learning areas.

Multiple previous investigations since 2010 have aimed to exploit information lost due to scattering media, like biological tissues, using mathematical approaches. Typically, this is done by applying optical operators such as linear scattering matrices, which help establish the input-output relations of photons as they scatter.

Professor Choi’s team from the IBS CMSD, primarily interested in this subject, has published several works combining hardware and software-based adaptive optics for tissue imaging. Some of their work has been displayed in new microscope types that can view through scattering media with high opacity, like mouse skulls, and perform in-depth 3D imaging of tissues.

However, when nonlinearity is introduced, things become exponentially more complex. When a scattering medium produces nonlinear signals, it cannot be simply depicted by a linear matrix as the superposition principle is breached. Moreover, assessing the nonlinear input-output characteristics becomes an immense challenge, making research in this area extremely demanding.

Demystifying Nonlinear Scattering Media

In this instance, Professor Choi’s team has achieved another groundbreaking discovery. They are the first to find that the input-output optical response of a nonlinear scattering medium can be expressed as a third-order tensor, as opposed to a linear matrix.

A third-order tensor is a mathematical entity used to illustrate relationships between three data sets. Simply put, it is an arrangement of numbers in a three-dimensional structure. Tensors are generalizations of scalars (0th-order tensors), vectors (1st-order tensors), and matrices (2nd-order tensors) and are frequently utilized in various mathematics, physics, and engineering fields to describe physical quantities and their relations.

The team demonstrated this concept by utilizing a medium made of barium titanate nanoparticles, which produce nonlinear second harmonic generation (SHG) signals due to barium titanate’s inherent noncentrosymmetric properties. The researchers came up with and experimentally validated a new theoretical framework that involves these cross-terms in a third-order tensor.

They accomplished this by measuring cross-terms by distinguishing the difference between the output electric fields produced when two input channels were activated at the same time, and when each channel was activated separately. This required an additional 1,176 measurements set by the possible combinations of two independent input channels, even with just 49 input channels.

Realizing Real-World Applications

The tensor derived from the nonlinear scattering medium has a higher rank than matrices of linear scattering media, indicating its potential as a scalable physical operator. The team showed this through the practical application of nonlinear optical encryption and all-optical logic gates.

The team successfully proved that nonlinear scattering media can be utilized for the optical encryption process. When specific image information is inputted into the media, the output second harmonic wave signals are displayed as random patterns, similar to a series of encryption processes.

By conducting an inverse operation of the third-order tensor representation of the second harmonic wave, the original input information can be retrieved through a decryption process. Using the inverse operation of the tensor input-output response, they decoded original signals from randomly encoded SHG signals, offering more security than standard optical encryption using linear scattering media.

Moreover, incorporating digital phase conjugation allowed the researchers to display all-optical AND logic gates that activate only when two specific input channels are simultaneously activated. This method presents potential benefits over silicon-based logic, including lower energy consumption and light-speed parallel processing capabilities.

This research is projected to pave new paths in the areas of optical computing and machine learning. “In the rapidly growing field of all-optical machine learning, nonlinear optical layers are critical in improving model performance. We are currently exploring how our research could be incorporated into this field,” stated Professor Choi.

This study was sponsored by the Institute for Basic Science.

Frequently Asked Questions (FAQs) about Optical Computing Breakthrough

More about Optical Computing Breakthrough

• IBS Center for Molecular Spectroscopy and Dynamics
• Optical Computing
• Machine Learning
• Nonlinear Scattering Media
• Nature Physics Journal