<p>The task of the project was to evaluate a differential phase shift keying demodulation technique by replacing a Mach-Zehnder interferometer receiver with an optical filter (Fiber Bragg Grating). Computer simulations were made with single optical transmission, multi channel systems and transmission with combined angle/intensity modulated optical signals. The simulations showed good results at both 10 and 40 Gbit/s. Laboratory experiments were made at 10 Gbit/s to verify the simulation results. It was found that the demodulation technique worked, but not with satisfactory experimental results. The work was performed at Eindhoven University of Technology, Holland, within the framework of the STOLAS project at the department of Electro-optical communication.</p>

The task of the project was to evaluate a differential phase shift keying demodulation technique by replacing a Mach-Zehnder interferometer receiver with an optical filter (Fiber Bragg Grating). Computer simulations were made with single optical transmission, multi channel systems and transmission with combined angle/intensity modulated optical signals. The simulations showed good results at both 10 and 40 Gbit/s. Laboratory experiments were made at 10 Gbit/s to verify the simulation results. It was found that the demodulation technique worked, but not with satisfactory experimental results. The work was performed at Eindhoven University of Technology, Holland, within the framework of the STOLAS project at the department of Electro-optical communication.

No description available.

The Internet has fundamentally changed the way of modern communication. Current trends indicate that high-capacity demands are not going to be saturated anytime soon. From Shannon's theory, we know that information capacity is a logarithmic function of signal-to-noise ratio (SNR), but a linear function of the number of dimensions. Ideally, we can increase the capacity by increasing the launch power, however, due to the nonlinear characteristics of silica optical fibers that imposes a constraint on the maximum achievable optical-signal-to-noise ratio (OSNR). So there exists a nonlinear capacity limit on the standard single mode fiber (SSMF). In order to satisfy never ending capacity demands, there are several attempts to employ additional degrees of freedom in transmission system, such as few-mode fibers (FMFs), which can dramatically improve the spectral efficiency. On the other hand, for the given physical links and network equipment, an effective solution to relax the OSNR requirement is based on forward error correction (FEC), as the response to the demands of high speed reliable transmission. In this dissertation, we first discuss the model of FMF with nonlinear effects considered. Secondly, we simulate the FMF based OFDM system with various compensation and modulation schemes. Thirdly, we propose tandem-turbo-product nonbinary byte-interleaved coded modulation (BICM) for next-generation high-speed optical transmission systems. Fourthly, we study the Q factor and mutual information as threshold in BICM scheme. Lastly, an experimental study of the limits of nonlinearity compensation with digital signal processing has been conducted.

IEEE has recently finished its ratification of the IEEE Standard 802.3ba in June 2010 which set the target Ethernet speed as 100 Gbps. The studies on the future trends of the ever-increasing demands for higher speed optical fiber communications show that there is no sign of decline in the demand. Constantly increasing internet traffic and the bandwidth-hungry multimedia services like HDTV, YouTube, voice-over-IP, etc. can be shown as the main driving forces. Indeed, the discussions over the future upgrades on the Ethernet speeds have already been initiated. It is predicted that the next upgrade will enable 400 Gbps Ethernet and the one after will be toward enabling the astounding 1 Tbps Ethernet.Although such high and ultra high transmission speeds are unprecedented over any transmission medium, the bottlenecks for achieving them over the optical fiber remains to be fundamental. At such high operating symbol rates, the signal impairments due to inter- and intra-channel fiber nonlinearities and polarization mode dispersion get exacerbated to the levels that cripple the high-fidelity communication over optical fibers. Therefore, efforts should be exerted to provide solutions that not only answer the need for high-speed transmission but also maintain low operating symbol rates.In this dissertation, we contribute to these efforts by proposing nonbinary-LDPC-coded modulation (NB-LDPC-CM) schemes as enabling technologies that can meet both the aforementioned goals. We show that our proposed NB-LDPC-CM schemes can outperform their prior-art, binary counterparts called bit-interleaved coded modulation (BI-LDPC-CM) schemes while attaining the same aggregate bit rates at a lower complexity and latency. We provide comprehensive analysis on the computational complexity of both schemes to justify our claims with solid evidence. We also compare the performances of both schemes by using amplified spontaneous emission (ASE) noise dominated optical fiber transmission and short to medium haul optical fiber transmission scenarios. Both applications show outstanding performances of NB-LDPC-CM schemes over the prior-art BI-LDPC-CM schemes with increasing gaps in coding gain as the transmission speeds increase. Furthermore, we present how a rate-adaptive NB-LDPC-CM can be employed to fully utilize the resources of a long haul optical transport network throughout its service time.

Amplificadores a fibra dopada com érbio de dupla passagem (DP-EDFA) contendo fibras compensadoras de dispersão (DCF) são estudados nesta tese. Os DP-EDFAs são uma alternativa para amplificadores de baixo custo e bom desempenho, sobretudo para utilização em redes metropolitanas. Devido a sua estrutura de dupla passagem, outras funções podem ser facilmente inseridas em seu circuito óptico, incrementando sua funcionalidade e reduzindo o custo envolvido. Considerando ainda que a compensação da dispersão é crítica para sistemas WDM de alta velocidade, um interessante tópico de investigação, que será bastante explorado nesta tese, refere-se a embutir uma fibra compensadora de dispersão dentro no circuito do DP-EDFA. Estudos experimentais são conduzidos, mostrando as vantagens e os problemas associados a algumas topologias de circuitos de DP-EDFAs. Um circuito original de DP-EDFA contendo DCF é proposto e caracterizado em termos de ganho e figura de ruído para diversas aplicações multicanal e multibanda dentro do espectro de transmissão da fibra óptica. Caracterizações sistêmicas em termos de taxa de erro de bit são também conduzidas / Double pass erbium doped fiber amplifiers (DP-EDFA) with an embedded dispersion compensation fiber are studied in this work. DP-EDFA is a cost-effective alternative for optical amplifiers, yet preserving good optical performance, mainly for use in metropolitan networks. Due to its double pass circuit, others functions beyond amplification can be inserted within the amplification circuit, increasing its capabilities and decreasing the involved cost. Considering that dispersion compensation is very critical for high-speed DWDM systems operating with standard single mode fiber, a timely topic of investigation concerns the embedding of the high insertion loss DCFs (dispersion compensating fibers) into optical amplifier configurations. Experimental studies are conducted first showing the advantages and impairments associated with DP-EDFAs circuits using DCFs. Next, a DP-EDFA proposed in this work is characterized in terms of gain and noise figure for several applications in multichannel and multiband network topologies. Systemic characterization in terms of bit error rate is also conducted, evaluating the performance of the proposed DP-EDFA using DCF

Esta dissertação apresenta uma análise realista de redes OCDMA (acesso múltiplo por divisão de código) baseadas em dois formatos de modulação distintos, qual sejam, OOK (chaveamento on-off) e PPM (modulação por posição de pulso). O desempenho desses formatos de modulação sob dois esquemas de codificação bidimensionais (2-D), isto é, pulsos únicos por linha (SPR) e múltiplos pulsos por linha (MPR) também são avaliados. Em ambos os casos, os efeitos simultâneos de muitos mecanismos dispersivos e ruídos (incluindo a interferência de acesso múltiplo, MAI) que prejudicam o desempenho geral do sistema foram também considerados. Outros mecanismos de degradação foram também incluídos, tais como o ruído de intensidade relativa (RIN) na etapa de transmissão, efeitos dispersivos da fibra óptica, como as dispersões de velocidade de grupo (GVD) e dos modos de polarização de primeira ordem (PMD), assim como os ruídos de batimento, APD, e térmicos na recepção. O efeito da GVD e PMD, bem como a influência dos ruídos, sobre o desempenho dos códigos SPR e MPR também são investigados. Outro efeito estudado nesta dissertação é a influência do fotodetetor APD sobre o ruído de batimento da rede OCDMA não-coerente. Para mitigar os ruídos do sistema e a BER, o algoritmo FEC RS (255,239) foi adotado em ambas as redes analisadas. Novas expressões para a taxa de erro de bit (BER) incluindo todos os ruídos e mecanismos de dispersão são apresentadas para os códigos SPR e MPR. Os resultados indicam que os esquemas de modulação OOK e PPM sem mecanismos adicionais de mitigação da MAI e de outros efeitos de ruídos não são suficientes para acomodar 32 usuários simultâneos em meio livre de erros (BER < \'10 POT.-12\'). Isto ocorre devido à alta BER na entrada do decodificador FEC, que afeta severamente a correção de erros na deteção. / This dissertation presents a comprehensive analysis of an OCDMA (optical code division multiple access) network based on two distinct modulation formats, namely, OOK (on-off keying) and PPM (pulse position modulation). We also investigate how each of these modulation formats performs under two distinct two dimensional (2-D) coding schemes, i.e., single-pulse per row (SPR) and multiple-pulse per row (MPR). For both cases, we have accounted for the simultaneous effect of many different dispersion and noise mechanisms (including multiple access interference, MAI) that impair the overall system performance. We have included the laser relative intensity noise (RIN) at the transmitter side, the fiber dispersive effects, such as group velocity dispersion (GVD) and first order polarization mode dispersion (PMD), as well as beat, APD, and thermal noises at the receiver side. The effect of GVD and PMD, as well as the influence of noises, on the performance of SPR and MPR codes is also investigated. Another effect studied in this dissertation is the influence of the avalanche photodetector (APD) photodetector on the beat noise of an incoherent OCDMA network. To mitigate systems noises and bit error rate (BER), we have adopted a FEC RS (255,239) algorithm in both networks investigated here. New expressions for the BER with all noises and dispersion mechanisms were also derived for the SPR and MPR code schemes. Results indicated that OOK and PPM modulation schemes without additional mechanisms to mitigate MAI and other noise effects are not sufficient to accommodate 32 simultaneous users in an error-free environment (BER < \'10 POT.-12\'). This occurs due to the already high BER at the FEC decoder input, which severely affects forward error corrections at the receiver side.

Electroabsorption is the physical phenomenon by which the absorption of light in a medium can be controlled by applying an electric field. The Quantum–Confined Stark Effect, which makes the absorption band–edge in quantum wells very field–dependent, together with the strong absorption peak provided by excitons, are the physical foundations for the success of electroabsorption modulators based on quantum well structures in telecommunication networks. This thesis describes the design and fabrication of surface–normal electroabsorption modulation devices. The techniques needed to understand the design and fabrication of surface–normal multiple quantum well optical modulators are introduced, as are the various characterisation techniques used during and after the fabrication. Devices for several types of applications have been designed, fabricated, characterised and in some cases integrated into optical systems: – Two–dimensional arrays of 128´128 pixel amplitude modulators grown on GaAs substrates have been fabricated and characterised. Speeds of up to 11700 frames per second were demonstrated, limited by the output electronics of the computer interface. – Large–area modulators grown on GaAs substrates for free–space optical communication were developed, with an active area of 2cm2 and a modulation speed of several megahertz. Contrast ratios up to 5:1 on full modulator areas were measured. Problems limiting the yield and modulation speed of such devices have been studied, and solutions to overcome them have been demonstrated. – Large–area devices grown on InP substrates for free–space optical communication have been developed. Contrast ratios of up to 2:1 for transmissive types have been demonstrated. – Devices consisting of two rows of pixels, grown on GaAs substrates, with an active area of 22mm´5mm, divided into 64 or 128 pixels per row have been developed. These amplitude modulation devices were designed for optical signal processing applications. – One variant of these optical signal processing devices was also characterised as a ternary, binary amplitude and binary phase modulator array. – The use of GaAs multiple quantum well optical modulators in a free–space optical retro–communication system has been studied. An opto–mechanical design for a modulating retro–reflector is described, allowing a large field of view in one direction using reflecting, resonant–cavity modulators for high contrast ratios. / QC 20100802

Among various planar lightwave circuits for multiplexing/demultiplexing in an optical communication system, etched diffraction gratings (EDGs) have shown great potential due to their compactness and high spectral finesse. Conventional numerical methods for grating simulation cannot be used to simulate an EDG demultiplexer of large size (in terms of the wavelength). In the present thesis, the polarization-dependent characteristics of an EDG demultiplexer are analyzed with a boundary element method (BEM) for both an echelle grating coated with a metal and a dielectric grating with total internal reflection (TIR) facets. For EDGs with metal-coated facets, we use a more effective method, namely, method of moments (MoM). Futhermore, a fast simulation method for EDGs with TIR facets is presented based on the Kirchhoff–Huygens principle and the Goos-Hänchen shift. This simple method has a good agreement with a BEM over a wide range of practical parameters of the device. Several novel designs are presented in order to improve the performances of EDGs. (1) By making some appropriate roughness on the surface of the shaded facets, the PDL of the demultiplexer can be effectively reduced over a large bandwith. (2) For EDGs based on Si nanowire structures, we compensate the polarization-dependent wavelength dispersion (PDλ) in the whole operational spectrum by introducing a polarization compensation area in its free propagation region. (3) An EDG demultiplexer with suppressed sidelobe is designed. The designed EDG demultiplexer can give a crosstalk as small as 50 dB in theory. (4) By chirping the diffraction order for each facet, we minimize the envelope intensity for the other adjacent diffraction orders to achieve a negligible return loss in a large spectral width. (5) A design for EDG demultiplexers is presented to obtain both large grating facets and a larger free spectral range (FSR) using the optimal chirped diffraction orders for different facets. The influences of the fabrication errors (e.g., rounded effect, surface roughness and point defect in the waveguide) on the performance (such as the insertion loss, the polarization dependent loss and the chromatic dispersion) of an EDG demultiplexer are also analyzed in detail. Silicon nanowire waveguides and related EDGs are studied. An EDG demultiplexer with 10 nm spacing is finally fabricated and characterized. / QC 20100910

Les télécommunications optiques en espace libre sont capables d'autoriser des débits de plusieurs dizaines de gigabits par seconde. Pour traiter ces débits en bénéficiant des techniques développées dans le cadre de l'optique fibrée, il est souhaitable d'injecter le faisceau reçu dans une fibre monomode. Ainsi, l'optique adaptative est proposée pour limiter les pertes d'injection induites par la dégradation de la qualité du faisceau propagé à travers la turbulence atmosphérique.Dans ce travail nous considérons un lien entre un satellite et une station de réception au sol. En élévation haute lorsque les perturbations d'amplitude peuvent être moyennées en augmentant la taille de la pupille la correction de la phase par optique adaptative (OA) classique peut être suffisante. Mes travaux dans ce cadre portent sur la mise en œuvre expérimentale de l'optimisation du couplage entre un système d'OA classique et une fibre monomode, en corrigeant les aberrations différentielles entre la voie de mesure de la surface d'onde et la voie d'injection du signal corrigé par OA. Le travail accompli dans l'objectif de cette mise en œuvre, se résume par les tâches suivantes : Définition d'une méthode d'optimisation automatique de l'injection par contrôle d'un seul miroir déformable pour la correction conjointe des perturbations rapides de la phase et des aberrations différentielles quasi-statiques ; Étude numérique de l'implantation de la méthode définie par simulation end-to-end ; Mise en œuvre de la méthode en laboratoire sur le banc BOA en modélisant la turbulence atmosphérique par des écrans de phase en rotation et validation du fonctionnement de l'optimisation de l'injection en boucle fermée ; Application de la méthode sur le banc ODISSEE à l'OCA puis fermeture de la boucle sur un signal optique émis par le satellite japonais SOCRATE.Par ailleurs, le lien satellite doit être opérationnel sur une durée maximale. Pour cela, le système doit fonctionner même dans le cas où le satellite est bas sur l'horizon et donc lorsque la portion d'atmosphère traversée par le faisceau est épaisse. Dans ces conditions particulières, les perturbations induites par les fluctuations de l'indice de l'air sont plus fortes que pour l'observation astronomique, donc la correction simultanée de la phase et de l'amplitude est indispensable pour atteindre le taux d'injection désiré (> 50 %). La deuxième partie de la thèse se place dans ce cadre et est consacrée à l'étude de l'implantation expérimentale d'une approche innovante de la correction de la phase et de l'amplitude fondée sur l'utilisation d'un Mach-Zehnder en optique intégrée. Le travail réalisé dans cette partie comporte les points suivants : Élaboration du principe de la correction par un Mach-Zehnder et proposition d'un dispositif généralisé à N voies ; Étude analytique de la correction par le dispositif proposé ; Développement d'un algorithme de contrôle fondé sur l'architecture proposée en optique intégrée ;• Validation numérique du fonctionnement par simulation end-to-end ; Conception d'un prototype pour la validation expérimentale du concept. / Free space optical communications are able to allow rates of several tens of gigabits/s. In order to process these flows and benefit from the techniques developed in the context of fiber optics, it is desirable to inject the received beam into a single mode fiber. In this case, adaptive optics is proposed to limit the injection losses induced by the degradation of the beam quality caused by the atmospheric turbulence.In this thesis, we consider a link between a satellite and a ground receiving station.For high elevations when amplitude perturbations can be averaged by increasing the size of the pupil and the correction of the phase is sufficient by conventional adaptive optics (AO), this work deals with the experimental implementation of the coupling optimization between a conventional AO and a single mode fiber, by correcting the non common path aberrations between the measurement path of the wave front and the injection path of the corrected signal. The accomplished work in this context is summarized by the following tasks: definition of an automatic injection optimization method by controlling a single deformable mirror for the parallel correction of fast phase disturbance and quasi-static differential aberrations; numerical study of the implementation of the proposed method defined end-to-end simulations ; implementation of the method in the laboratory on the BOA bench, modeling the atmospheric turbulence by rotating phase screens, and validation of the injection optimization in closed loop ; application of the method on the ODISSEE bench at the OCA and closing of the loop on a laser signal emitted by the Japanese satellite SOCRATE.Besides, the link must be operational for a maximum duration. Indeed, the system must operate even when the satellite is low on the horizon and, therefore, when the crossed portion of atmosphere is thick. In these particular conditions, the disturbances induced by the air index fluctuations are stronger than for astronomical observations, so that, the simultaneous correction of the phase and the amplitude is essential to reach the desired injection ratio (> 50 %). Based on this analysis, the second part of this thesis focuses on the study of the experimental implementation of an innovative approach of a phase and amplitude correction based on using an integrated Mach-Zehnder. The work done in this part consists in the following points : elaboration of correction principle by a Mach-Zehnder and proposition of a generalized device with N channels ; theoretical analysis of the superposition by the proposed device ; development of a control algorithm based on the architecture proposed in integrated optics ; numerical validation of operation by end-to-end simulations ; design of the prototype for the experimental validation of the concept in the laboratory.