Search

This thesis lies within the field of unconventional computing, and in particular focuses on physical systems whose dynamics can usefully process information.Such platforms may not outperform an electronic processor but may offer lower power consumption and better integration with sensors, and consequently may be useful for edge computing applications.Here, we study photonic systems, leveraging the relative maturity of the field due to the growth of telecommunication-related industries.Specifically, we explore how the wavelength degree of freedom of the light can be exploited to implement frequency multiplexed Artificial Neural Network (ANN) algorithms.Among many kinds of ANN algorithms, we focus on randomized ANNs. These networks are mostly composed of randomly selected and untrained connections, and only the output connections are tuned. Randomized ANNs offer interesting advantages, such as the possibility of low-cost training (both in terms of time and power) and the simplicity of implementation in physical computers. We describe the implementation of three randomized ANN algorithms: Extreme Learning Machine (ELM), Reservoir Computer (RC), and Deep Reservoir Computer (Deep-RC).We describe a variety of schemes where artificial neurons are encoded in the lines of an optical frequency comb.The mixing of information, necessary for machine learning, happens either using periodic phase modulation or using Kerr nonlinearity.Output weights can be applied optically by using programmable spectral filters.Each scheme is experimentally demonstrated in a fiber-based setup and evaluated on standard machine learning benchmark tasks.Benchmark results are in line with the performance reported on similar photonic-based systems and equivalent software implementations. We demonstrate that parallel operations can be executed on a single hardware.Parallel-running computations can be concatenated, thus constituting the layers of a Deep-RC.To the best of our knowledge, we report the firs, Cette thèse s'inscrit dans le domaine de l'informatique non conventionnelle et se concentre sur les systèmes physiques dont la dynamique peut être utilisée pour traiter l'information.Ces plateformes ne sont pas forcément plus performantes qu'un processeur électronique, mais peuvent offrir une consommation d'énergie plus faible et une meilleure intégration avec les capteurs, et peuvent donc être utiles pour les applications de edge computing.Nous étudions ici des systèmes photoniques, en tirant parti de la maturité relative du domaine due à la croissance de l’industries des télécommunications.Plus précisément, nous étudions comment le degré de liberté de la longueur d'onde de la lumière peut être exploité pour mettre en œuvre des algorithmes de réseaux neuronaux artificiels multiplexés en fréquence.Parmi les nombreux types d'algorithmes de réseaux neuronaux artificiels, nous nous concentrons sur les architectures randomisées.Ces réseaux sont principalement composés de connexions sélectionnées au hasard et non entraînées, et seules les connexions de sortie sont ajustées.Ces architectures randomisées offrent des avantages intéressants, tels que la possibilité d'un entraînement peu coûteux (à la fois en termes de temps et d'énergie) et la simplicité de mise en œuvre dans les ordinateurs physiques.Nous décrivons la mise en œuvre de trois algorithmes :l’Extreme Learning Machine (ELM), le Reservoir Computer (RC) et le Reservoir Computer Profond (Deep-RC).Nous décrivons une variété de schémas dans lesquels les neurones artificiels sont encodés dans les lignes d'un peigne de fréquence optique.Le mélange des informations, nécessaire à l'apprentissage automatique, se fait soit par modulation de phase périodique, soit par non-linéarité de Kerr.Les poids de sortie peuvent être appliqués optiquement à l'aide de filtres spectraux programmables. Chaque schéma fait l'objet d'une démonstration expérimentale en fibre optique et est évalué sur des tâches de référence standard en matièr, Doctorat en Sciences, info:eu-repo/semantics/nonPublished

We demonstrate a nonlinear photomodulation spectroscopy method to image the mode profile of a high-Q photonic crystal resonator (PhCR). This far-field imaging method is suitable for ultrahigh-Q cavities which we demonstrate on a Q=619000 PhCR. We scan the PhCR surface with a 405-nm pump beam that modulates the refractive index by local thermal tuning, while probing the response of the resonance. We enhance resolution by probing at high power, using the thermo-optical nonlinearity of the PhCR. Spatial resolution of the thermo-optical effect is typically constrained by the broad thermal profile of the optical pump. Here we go beyond the thermal limit and show that we can approach the diffraction limit of the pump light. This is due to free carrier absorption that heats up the PhCR only when there is overlap between the optical pump spot and the optical mode profile. This is supported with a thermo-optical model that reproduces the high-resolution mode mapping. Results reveal that the observed enhanced resolution is reached for surprisingly low carrier density.

Stochastic computing allows a drastic reduction in hardware complexity using serial processing of bit streams. While the induced high computing latency can be overcome using integrated optics technology, the design of realistic optical stochastic computing architectures calls for energy efficient switching devices. Photonics Crystal (PhC) nanocavities are $\mu m^2$ scale devices offering 100fJ switching operation under picoseconds-scale switching speed. Fabrication process allows controlling the Quality factor of each nanocavity resonance, leading to opportunities to implement architectures involving cascaded gates and multi-wavelength signaling. In this report, we investigate the design of cascaded gates architecture using nanocavities in the context of stochastic computing. We propose a transmission model considering key nanocavity device parameters, such as Quality factors, resonance wavelength and switching efficiency. The model is calibrated with experimental measurements. We propose the design of XOR gate and multiplexer. We illustrate the use of the gates to design an edge detection filter. System-level exploration of laser power, bit-stream length and bit-error rate is carried out for the processing of gray-scale images. The results show that the proposed architecture leads to 8.5nJ/pixel energy consumption and 512ns/pixel processing time.

We theoretically and experimentally investigate type II second harmonic generation in III-V-on-insulator wire waveguides. We show that the propagation direction plays a crucial role and that longitudinal field components can be leveraged for robust and efficient conversion. We predict that the maximum theoretical conversion is larger than that of type I second harmonic generation for similar waveguide dimensions and reach an experimental conversion efficiency of 12%/W, limited by the propagation loss., SCOPUS: ar.j, info:eu-repo/semantics/published

We investigate second-harmonic generation in III-V semiconductor wire waveguides aligned with a crystallographic axis. In this direction, because of the single nonzero tensor element coming from the zinc-blende symmetry of III-V semiconductors, only frequency conversion by mixing with the longitudinal components of the optical fields is allowed. We experimentally study the impact of the propagation direction on the conversion efficiency and confirm the role played by the longitudinal components through the excitation of an antisymmetric second-harmonic higher-order mode., SCOPUS: ar.j, info:eu-repo/semantics/published

info:eu-repo/semantics/published

Resonant cavities with high quality factor and small mode volume provide crucial enhancement of light-matter interactions in nanophotonic devices that transport and process classical and quantum information. The production of functional circuits containing many such cavities remains a major challenge, as inevitable imperfections in the fabrication detune the cavities, which strongly affects functionality such as transmission. In photonic crystal waveguides, intrinsic disorder gives rise to high-Q localized resonances through Anderson localization; however their location and resonance frequencies are completely random, which hampers functionality. We present an adaptive holographic method to gain reversible control on these randomly localized modes by locally modifying the refractive index. We show that our method can dynamically form or break highly transmitting necklace states, which is an essential step toward photonic-crystal-based quantum networks and signal processing circuits, as well as slow light applications and fundamental physics.

info:eu-repo/semantics/published

We propose high index contrast InGaP photonic wires as a platform for the integration of nonlinear optical functions in the telecom wavelength window. We characterize the linear and nonlinear properties of these waveguide structures. Waveguides with a linear loss of 12 dB/cm and which are coupled to a single mode fiber through gratings with a -7.5 dB coupling loss are realized. From four wave mixing experiments, we extract the real part of the nonlinear parameter γ to be 475 ± 50 W-1m-1 and from nonlinear transmission measurements we infer the absence of two-photon absorption and measure a three-photon absorption coefficient of (2.5 ± 0.5) × 10-2 cm3GW-2., SCOPUS: ar.j, info:eu-repo/semantics/published

With the help of the strong nonlinear optical response in silicon nanophotonic wire waveguides complex sources and devices can be integrated on a chip. We demonstrate mid-infrared wavelength translators and frequency combs Furthermore, we open up the possibility for the integration of nonlinear optical functions at telecom wavelengths by exploiting more exotic nonlinear interactions in silicon waveguides as well as new waveguide platforms. © 2015 IEEE., info:eu-repo/semantics/published

We demonstrate the generation of an octave-spanning supercontinuum in InGaP membrane waveguides on a silicon substrate pumped by a 1550-nm femtosecond source. The broadband nature of the supercontinuum in these dispersion-engineered high-index-contrast waveguides is enabled by dispersive wave generation on both sides of the pump as well as by the low nonlinear losses inherent to the material. We also measure the coherence properties of the output spectra close to the pump wavelength and find that the supercontinuum is highly coherent at least in this wavelength range., SCOPUS: ar.j, info:eu-repo/semantics/published

We present theory and measurements ofdisorder-induced losses for low loss 1.5 mmlong slow light photonic crystal waveguides. A recent class of dispersion engineered waveguides increases the bandwidth of slow light and shows lower propagation losses for the same group index. Our theory and experiments explain how Bloch mode engineering can substantially reduce scattering losses for the same slow light group velocity regime.

The dynamical properties of an InP photonic crystal nanocavity are experimentally investigated using pump-probe techniques and compared to simulations based on coupled-mode theory. Excellent agreement between experimental results and simulations is obtained when employing a rate equation model containing three time constants, that we interpret as the effects of fast carrier diffusion from an initially localized carrier distribution and the slower effects of surface recombination and bulk recombination. The variation of the time constants with parameters characterizing the nanocavity structure is investigated. The model is further extended to evaluate the importance of the fast and slow carrier relaxation processes in relation to patterning effects in the device, as exemplified by the case of all-optical wavelength conversion.

The availability of a tunable delay line with a chip-size footprint is a crucial step towards the full implementation of integrated microwave photonic signal processors. Achieving a large and tunable group delay on a millimetre-sized chip is not trivial. Slow light concepts are an appropriate solution, if propagation losses are kept acceptable. Here we use a low-loss 1.5 mm-long photonic crystal waveguide to demonstrate both notch and band-pass microwave filters that can be tuned over the 0 50-GHz spectral band. The waveguide is capable of generating a controllable delay with limited signal attenuation (total insertion loss below 10 dB when the delay is below 70 ps) and degradation. Owing to the very small footprint of the delay line, a fully integrated device is feasible, also featuring more complex and elaborate filter functions.

The availability of a tunable delay line with a chip-size footprint is a crucial step towards the full implementation of integrated microwave photonic signal processors. Achieving a large and tunable group delay on a millimetre-sized chip is not trivial. Slow light concepts are an appropriate solution, if propagation losses are kept acceptable. Here we use a low-loss 1.5 mm-long photonic crystal waveguide to demonstrate both notch and band-pass microwave filters that can be tuned over the 0 50-GHz spectral band. The waveguide is capable of generating a controllable delay with limited signal attenuation (total insertion loss below 10 dB when the delay is below 70 ps) and degradation. Owing to the very small footprint of the delay line, a fully integrated device is feasible, also featuring more complex and elaborate filter functions.

The OPTHER (Optically Driven THz amplifier) project supported by the European Commission within the Seventh Framework Program (FP7) represents the first joint European attempt to realize vacuum electron devices in THz range. The target of the project was to design and realize the first 1 THz vacuum tube amplifier. The challenges of the presented task and the innovative solutions adopted established a new level of knowledge in the field. The main aspects of the OPTHER project are described, focusing on challenges and adopted innovative solutions.

The OPTHER (Optically Driven THz amplifier) project supported by the European Commission within the Seventh Framework Program (FP7) represents the first joint European attempt to realize vacuum electron devices in THz range. The target of the project was to design and realize the first 1 THz vacuum tube amplifier. The challenges of the presented task and the innovative solutions adopted established a new level of knowledge in the field. The main aspects of the OPTHER project are described, focusing on challenges and adopted innovative solutions.

Photonic crystals will play a fundamental role in the future of optical communications. The relevance of the numerical modeling for the success of this technology is assessed by using some examples concerning the experience of the COPERNICUS Project.