Abstract

For photonic devices, structural disorder and light scattering have long been considered annoying and detrimental features that were best avoided or minimized. This review shows that disorder and complexity can be harnessed for photonic device applications. Compared to ordered systems, disordered systems provide much more possibilities and diverse optical responses. They have been used to create physical unclonable functions for secret key generation, and more recently for random projection, high-dimensional matrix multiplication, and reservoir computing. Incorporating structural disorder enables novel devices with unique functionalities as well as multi-functionality. A random system can function as an optical lens, a spectrometer, a polarimeter, and a radio frequency receiver. It is also employed for optical pulse measurement and full-field recovery. Multi-functional disordered photonic devices have been developed for hyperspectral imaging, spatial, and spectral polarimetry. In addition to passive devices, structural disorder has been incorporated to active devices. One prominent example is the random laser, which enables speckle-free imaging, super-resolution spectroscopy, broad tunability of high-power fiber laser, and suppression of lasing instabilities. Disordered devices have low fabrication costs, and their combination with advanced computational techniques may lead to a paradigm shift in photonics and optical engineering.

Highlights

  • While most photonic devices are designed to have regular, ordered structures, disordered structures provide much more possibilities

  • This review shows that disorder and complexity can be harnessed for photonic device applications

  • This review aims to illustrate why structural disorder is useful for photonic device applications, and how it has been incorporated to both passive and active devices

Read more

Summary

INTRODUCTION

While most photonic devices are designed to have regular, ordered structures, disordered structures provide much more possibilities. An ordered structure has only a few spatial frequencies [Fig. 1(c)], but a disordered structure contains many more spatial frequencies [Fig. 1(d)], providing a huge parameter space for device design. Disordered structures are easy to fabricate and have low costs, while providing diverse responses to external signals. As photonic devices, their functionalities often rely on post-processing of their responses using computational algorithms, including compressive sensing and machine learning. Their functionalities often rely on post-processing of their responses using computational algorithms, including compressive sensing and machine learning Such algorithms have advanced rapidly in recent years; physical size and cost of processors have fallen sharply. The emergence of disordered photonic devices enables a paradigm shift of device design from hardware to software, which will dramatically lower the cost

Motivation
Random spectrometers
Multi-functional photonic devices
Suppression of coherent imaging artifacts
Photonic systems with disorder
Characteristic lengths of optical scattering
Light diffusion and localization
Multimode optical fiber
Spectral correlation function
Optical memory effect
PHOTONIC PHYSICAL UNCLONABLE FUNCTIONS
Introduction of PUFs
Linear scattering PUFs
VIII D VIII E VIII F
Reconfigurable optical PUFs
Nonlinear optical PUFs
Quantum readout
Strong PUFs
Weak PUFs and unique objects
Transfer function
Coherent imaging
Incoherent imaging
Compressive sensing and machine learning
Correlation-based imaging
Cross-correlation
Auto-correlation
Holographic imagers and wavefront sensors
Multimode fiber endoscopy
SPECKLE-BASED SPECTROMETERS
Spectrometer parameters
Random optical diffraction
Multiple light scattering
Multimode waveguide dispersion
Computational spectrometers
SNAPSHOT HYPERSPECTRAL IMAGERS
SCATTERING POLARIMETERS
Spatial polarimetry
Spectral polarimetry
TEMPORAL MEASUREMENT
Compressive radio-frequency receivers
Single-shot full-field measurement of optical pulses
Parallel temporal ghost-imaging
Nonlinear speckle and deep learning
VIII. RANDOM LASERS
Introduction of random laser
Spatial coherence engineering
Speckle-free imaging
Super-resolution spectroscopy
High power with high efficiency
Suppression of lasing instability
Disorder-enhanced optical sensors
Random projection
Reservoir computing
Reconfigurable operations
Engineering disorder
Findings
Idet Idif Iin IMMF Iout
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call