Four commercial single mode silica fibers were subjected to aging testing in hot water at different temperatures and durations. Static fatigue testing have showed an oscillating behaviour of fiber failure strength versus the aging duration, that may be explained through the structural relaxation phenomenon at the glass-polymer coating interface. For several aging durations, fibers appeared more resistant under static stress than the non-aged ones. The fiber strength measured dynamically, using a two point bending test, slightly decreased in the same aging conditions. Dynamic fatigue testing of as-aged fibers under permanent stress revealed an overall decrease of failure stress and a change of Weibull distribution towards a bi-modal dispersion. As-aged fibers presented a permanent curvature following aging treatment and drying.

If microstructured optical fibers are to find widespread use in photonics technology, they will have to be easily cleavable using mechanical cleavers, since more sophisticated cleaving techniques add complexity. Conventional mechanical cleavers are the preferred laboratory and production tools because they are both simpler to use and more time- and cost-effective compared to techniques such as laser cutting. When designing complex microstructured fibers (MSF) with exciting novel optical characteristics, it is therefore important to favor those geometries that allow high-quality cleavage using standard mechanical cleavers. In this paper, we present an analytical model for fracture propagation during the cleaving process in a complex MSF. The model is based on experimental observations. Three samples of high air-fraction double-clad MSFs were used. Although that they all feature the same structural profiles (but differing in certain geometrical dimensions), they give totally different cleavage patterns. The cleaved surfaces were studied and analyzed. Analysis of the cleaved surfaces allowed to establish a criterion for smooth fracture propagation in a high air-fraction double clad MSF and to suggest a novel design approach for these specific structures.

High-speed transmission optical devices often use a single-mode fiber as receiver. The fiber must be accurately aligned over five degrees of freedom with respect to an optical field. In previous works, it has been demonstrated that the optical coupling between axisymmetric Gaussian beams has distinctive parabolic, hyperbolic, and linear like characteristics that may be advantageously used to design automated alignment strategies. These properties were proven to exist experimentally when two single-mode fibers are used as receiver and emitter. This paper presents an experimental investigation of these properties for a practical optical coupling situation: the alignment of a receiving single-mode fiber with an optical device. For the purpose of this investigation, the optical device is comprised of an emitting fiber and two lenses mounted in series in order to form a converging optical field at the output of the second lense. A five axis nanopositionning system is used to move the receiving fiber relatively to the device. Even though optical fields are not exactly Gaussian, experiments demonstrate the existence of the properties within a practical range of interest for single-mode device-to-fiber alignment automation. These properties of the coupled optical power provide a strong basis to develop model-based algorithms for axisymmetric single-mode device-to-fiber alignment automation.

The paper summarizes the research of the iNEMI (International Manufacturing Initiatives) project on Fiber Connector Endface Specifications focused on the development of the cleanliness specification of the receptacle modules. It will discuss (1) the critical parameters of Small Form-Factor (SFF&SFP) modules sensitive to the contamination, (2) the experimental contamination techniques, (3) the test and measurement methodology, as well as (4) the Gage R & R (Repeatability & Reproducibility) studies. We choose one of the standard OC-48, SFF modules manufactured by Sumitomo Electric as the experimental vehicle for the project. The influence of the SiO2 based dust with the particle size under 149 um on optical performance of SFF module has been investigated. The Optical Return Loss (ORL) has been identified as the most sensitive parameter to the contamination for the both transmitter and receiver optical endface. In a case of changes of the transmitter ORL by 20-25 dB due to the application of the contamination, the optical power and spectral width didn't change, and the pulse mask was reduced only by 2%. The contamination near the core area caused changes of the ORL by 20-30dB, and high standard deviation of ORL (0.5-2.1 dB). The Gage R & R study was focused on investigation of equipment repeatability, mating repeatability, and operator dependence for ORL measurement of clean and contaminated optical endface. The proposed experimental methodology was able to achieve all Gage R & R parameters within the required limits.

In this paper, the polarization characteristics of fiber Bragg gratings written into spun fibers with certain intrinsic birefringence and short spinning periods are studied experimentally. In terms of the Stokes parameters, the evolutions of states of polarization (SOP) in spun fiber Bragg gratings are investigated. The polarization-mode dispersion (PMD) and polarization dependent loss (PDL) are examined in both the transmission and reflection manners. The spun fiber Bragg gratings exhibit distinctive properties from those in the conventional single-mode fibers. In particular, they always exhibit multiple side peaks near the Bragg peaks with certain wavelength intervals in their reflection spectra. It is further shown that these reflection peaks have different polarization dependences on input SOP. The observed polarization dependences in the experiments are then confirmed in the simulation, and potential applications are discussed.

A three-fiber probe is fabricated for the measurement of fluorescent light emitted from a flat and tilted membrane cast on a glass substrate. With two identical photonic crystal fibers (PCFs) as the receiving fibers, we demonstrate that a segment of pure glass rod formed at the front end of a large-core PCF has an enhanced capability of collecting fluorescent light. The percentage of this increase depends on the distance between the membrane and the fiber probe. An initial theoretical investigation reveals that a longer fused segment can shift the light reception cone towards the symmetric axis of the illuminating fiber where the maximum energy locates. Higher emission is expected in this area and would be able to reach this shifted reception cone. This area is a dead zone for the PCF without or with a shorter fused segment. The possible applications of this fused glass segment with a proper length are highlighted.

In this paper, the effects of the asymmetric refractive index change profile on the reflection spectra of multimode fiber Bragg gratings (MMFBGs) are experimentally and theoretically investigated. Different guided modes in a multimode fiber (MMF) are coupled with each other and the sub-reflection peaks are generated if the refractive index change profile of the MMFBG is asymmetric in the cross section of the fiber core. It is found that by increasing the UV exposure, the reflectivities of the main reflection peaks are decreased, the reflectivities of the sub-peaks are increased, even higher than the main reflection peaks, which indicate the grating becomes more asymmetric with increasing the UV exposure. It is further shown that an MMFBG is much more sensitive to the polarization state of the injected light compared with an SMFBG. The numerical simulations are carried out to explain the excitation condition dependence of the reflection spectra of MMFBG, which agree well with the experimental results.

This work demonstrates the feasibility of designing and photoimprinting fiber Bragg gratings with grating periods at least twice longer than conventional short-period fiber Bragg gratings but with identical performance, functionality and applications. A theoretical analysis has been conducted based on a new complete and coherent photosensitivity model for Ge-doped silica fibers. This new photosensitivity model deals in depth with the issue of microscopic mechanisms for grating formation, giving as main result the understanding of the photoinduced refractive index generating process.

The characterization of material and structural properties is essential in the development of high-performance optoelectronics devices. The gain and spontaneous emission of semiconductor emitters are intrinsically related, and knowing one determines the other. In this paper, we report on a comparison between the measured and calculated spontaneous emission spectra of complex semiconductor structures that were developed in our laboratory. Transversely emitted spontaneous emission spectra over a wide range of carrier densities have been obtained for GRIN-SCH-MQW In_xGa_1-xAs_yP_1-y structures consisting of three tensile and three compressive wells. Information from these measurements and materials parameters were used to estimate carrier density for each well and subsequently used in the calculation of the emission spectra. The theoretical results were obtained by calculating the spontaneous emission rate for each well independently and then by summing over the six wells. We first calculate the band structure from a 6x6 Luttinger-Kohn Hamiltonian and find the spontaneous emission rate using the carrier density obtained from experimental measurements. A comparison between the Markovian (Lorentzian) and non-Markovian (Gaussian) line shape functions is established, considering the bandgap renormalization. We show that the Gaussian broadening function gives better agreement with the experimental data.

InSb has been intensively studied in decades and widely used for fabricating high-performance devices because of its good chemical stability, low effective mass, high electron and hole mobility, and narrow band gap. The most important device applications for InSb are in thermal image sensing in the mid-infrared (3-5 μm) spectral range. The industry standard for fabricating InSb-based focal plane arrays for thermal imaging is based on indium bump technology to interconnect the InSb array to a Si-based readout integrated circuit chip. This hybridization is a "one-piece-at-a-time" process and thus time-consuming and costly. An alternative approach is to employ a device that up-converts mid-infrared light to a wavelength below 1 μm, which can then efficiently be detected by Si charged coupled devices. We reported herein such a mid-infrared optical up-converter based on InSb using wafer fusion technology. The up-conversion device consists of an InSb p⁺nn⁺ photodiode and a GaAs/AlGaAs LED, which were grown separately and wafer-bonded together. Experimental results demonstrated mid-infrared to 0.84 μm up-conversion operation at 77K. The measured LED external efficiency and photodiode responsivity show that an external up-conversion efficiency of 0.093 W/W was obtained. Effects of electrical gain and photon recycling inside this integrated device are discussed.

We present all-fiber double-clad Yb3+-doped fiber laser capable to switch between three different wavelengths. The fiber laser device consists of an array of three fiber Bragg gratings at 1064, 1080, and 1096 nm, spliced to one end of the double-clad Ytterbium-doped fiber, thus forming three laser cavities that share the same gain medium. The selection of a specific wavelength is realized by induced bend loss in the sections between gratings, thus allowing the control of the feedback at each specific wavelength. The device can operate separately at 1064, 1084 and at 1096 nm, with slope efficiencies at these three wavelengths of the order of 48% with respect to the launched pump power.

The key point to scale the output power of diode-pumped solid-sate (DPSS) laser is to solve the thermally induced problems such as fracture of the material by thermal stress, degradation of the beam quality and efficiency by thermally induced birefringence and aberration of the thermal lens, etc. For end-pumped solid state lasers, the gain medium can be constructed in a format of a thin disk or a composite rod to scale the output power. This concept has been successfully used to scale DPSS laser output powers by 1~2 order, depending on the laser material and the beam quality of the output. In a thin disk laser, pump induced heat flows predominantly along the thickness of the laser disk and the thermal lens is eliminated to first order. However, a conventional thin disk laser requires a complex and expensive multipass pump setup to maximize pump absorption for the thinnest crystal possible to minimize the residual thermal lens. Alternatively, using a composite rod in a conventional end-pumped DPSS laser elevates the maximum allowable pump power by ~50%, since the interface between the doped and undoped region of the gain medium provides a heat buffering effect and the maximum thermal stress is reduced. Our anvil-cell disk laser, which clamps the gain medium between the heat sink and a sapphire window, combines the benefits of both the thin-disk laser and lasers using composite rods but with the ability to further optimize material properties. In addition, the portion of thermal lens due to bulge of the gain medium can be compensated by pressure tuning. The complexity and cost on pump setup can be greatly reduced with this relatively simple design. In this work we demonstrated a reliable high power Nd:YVO₄ laser which delivered 26.2 W of laser output at M²=3. Results of intracavity frequency doubling with this laser are also reported.

We have demonstrated an electroabsorption modulator and semiconductor optical amplifier monolithically integrated with novel dual-waveguide spot-size converters (SSC) at the input and output ports for low-loss coupling to planar light-guide circuit silica waveguide or cleaved single-mode optical fiber. The device was fabricated by means of selective-area MOVPE growth, quantum well intermixing and asymmetric twin waveguide technologies with only a three steps low-pressure MOVPE growth. For the device structure, in SOA/EAM section, double ridge structure was employed to reduce the EAM capacitances and enable high bit-rate operation. In the SSC sections, buried ridge structure (BRS) were incorporated. Such combination of ridge, ATG and BRS structure is reported for the first time in which it can take advantage of easy processing of ridge structure and the excellent mode characteristic of BRS. At the wavelength range of 1550~1600nm, lossless operation with extinction ratios of 25 dB dc and more than 10 GHz 3-dB bandwidth is successfully achieved. The beam divergence angles of the input and output ports of the device are as small as 8.0°×12.6°, resulting in 3.0 dB coupling loss with cleaved single-mode optical fiber.

We report simulation results for a directional coupler between silicon waveguides in different layers of a three-dimensional (3D) optical circuit. The coupling length is 1.4 mm. The device is manufacturable using standard CMOS technology provided individual waveguide layers can be vertically stacked. In simulations of coupling efficiency the design exhibits negligible loss with respect to translational and rotational misalignments of up to 0.5 μm. Efficiency degradation is less than 5% for etch depth and waveguide width variations of 0.4 μm, and less than 1 dB over the range of standard lithographic tolerances for variations from layer to layer in feature width, depth, and alignment.

Photoresponsive nanocomposite organically modified silica (ORMOSIL) films were prepared by solution sol-gel processing of organically modified silicon alkoxide compounds. Waveguiding at 488 nm proceeds with simultaneous self-inscription and permanently inscribed waveguides can be revealed by wet etching. Under certain conditions, self-inscription becomes chaotic, and filamentation is observed. A counterpropagating beam set-up allows simple optical devices like crosses and y-junctions to be created. Soliton-like behavior is exhibited at low laser power where couterpropagating self-inscribed beams undergo mutual trapping. The ORMOSIL films exhibit interesting patterns which may be associated with the relief of stress in the films. These patterns can be controlled to some extent by depositing self-inscribed features in the glassy medium.

This paper presents photo-induced signal taps for power monitors in planar lightwave circuits. Both periodic perturbation and non-periodic perturbation structures are discussed. The numerical models based on the volume current method are used to simulate the interaction of guided light with photo-induced periodic and non-periodic perturbations. The results are compared with those from a commercially available FDTD simulation tool. The optimized designs for the minimized polarization dependent loss taps are presented. Contrary to the power monitors currently used in planar lightwave circuits, which are based on directional coupler and deep etch technology, we propose the signal taps through photosensitive processes using laser light. Our approach is more cost efficient, flexible and reliable.

In this presentation, first we investigate the diffusion process analytically and numerically. Some important fabrication and design parameters are abstracted and used for the next level simulation. Then we applied the efficient method to analyze the modal properties of the diffused waveguides. The accuracy and the scope of validity for the analytical methods are investigated and the modal properties directly link to the fabrication parameters. Based on modal properties of optimized waveguides, the device performances of the related devices can be easily obtained. The relation between the waveguide modal performance and design parameters is built and the possible optimization methods can be applied. Finally, some related issues such as the bending loss and the coupling loss with the standard fiber are discussed.

As the development of the technology, fiber-to-the-home (FTTH) becomes a feasible solution to meet the increasing demand on bandwidth. Due to the massive number of end users, cheap and reliable components become the bottleneck to deploy the new technology. Triplexer is one of the key components in the FTTH and is used by every end user. Currently, the available triplexers are either based on bulk optics or fiber optics with large size and high price due to manual labor involved. Planar lightwave circuit (PLC) is a possible technology for massive production and cost reduction. However, it is very challenging to design such bi-directional triplexer on PLC. The first challenge is that three channels, at λ=1310nm, 1490nm, and 1555nm, are separated unevenly over a very large wavelength range; Secondly, the bandwidths of the three channels, Δλ=100nm, 20nm, and 10nm, are very different. In the paper, we proposed a novel design by combining both coarse WDM and dense WDM. In the design, a multi-mode interference (MMI) device is used for coarse WDM to separate the 1310nm from the other two channels. The dense WDM for the remaining two channels is performed by an array waveguide gratings (AWG). The MMI and AWG are built on the same wafer with monolithic integration. Initial simulation results show it is a very promising device.

The theoretical analysis of a flat-top AWG with multimode interference (MMI) power splitter at the input end is done. Several corrections to the former papers are made. For example, the field coupled into the arrayed waveguides is not treated with approximation, and domain of integration for coupling coefficient of an arbitrary waveguide in the array is chosen to be the waveguide spacing, which is coincident with the real optical propagation in AWG. Expressions for spatial distribution of field are obtained at the interfaces between waveguides, while an analytic expression for each channel response of AWG is derived. All of the expressions obtained are the functions of AWG configuration parameters. The output response of each channel is obtained by the analytic method much more quickly than by beam propagation method (BPM), and the period of optimization design of AWG is shortened greatly. Finally, a design example of flat-top AWG with good performance is presented, and its important specifications are calculated.

The coupling efficiency of a pigtailed optical fiber lens is often a critical parameter in the manufacture of optical devices. For example, efficient coupling from high power semiconductor lasers to optical fiber pigtails with fiber lenses is of great importance as the uncoupled power can affect the lifetime of the device through thermal degradation of the pigtail or its welding point; for optical amplifiers, the coupling efficiency determines the overall gain. Generally, mass-produced optical fiber lenses do not have a consistent enough quality and it is often necessary to discard lenses that exhibit poor coupling. We previously demonstrated a method for digitally modifying optical fiber lenses. We have now developed a scheme whereby we are able to produce high quality optical fibre lenses with a given lens radius and then to iteratively alter it to achieve the required divergence and spot size in a repeatable fashion for efficient coupling to a variety of devices such as semiconductor optical amplifiers, lasers and wave-guides. We have consistently demonstrated 75% coupling efficiency to angled facet semiconductor optical amplifiers at 1550nm with only three-axis adjustment, and believe the true coupling efficiency to be close to 90%. Our powerful scheme allows the use of nearly all of the lenses for a given application.

Evaporated films of the azo-material DR 1 have been investigated. In the as-grown state partly crystalline films with low transmittance are obtained. Using homogeneous exposure transparent regions may be formed. The recording of holographic gratings in thin films (< 1 μm) of the azo dye is investigated for the case of more-dimensional light-intensity patterns. The mechanism of the photo-isomerisation of the azo-compounds is used to form dual gratings with a refractive index grating and a surface-relief grating. The grating-formation is investigated in case of 1D-gratings first. The time dependent diffraction efficiency is discussed in a model of 2 processes with different time-constants. A material transport process is involved in the formation of relief patterns. The enhancement of the modulation-depth of the surface-relief gratings is investigated for the application of a Corona discharge and a thermal treatment after the holographic recording. 2D-gratings are formed using either a 3-beam holographic set-up or a consecutive method. The resulting light patterns are simulated. Diffraction patterns and AFM-measurements are used to confirm these simulated structures. The modulation of the surface-relief gratings can be enhanced by thermal treatment after the holographic recording.

This paper describes a novel A/D converter called "Binary Delta-Sigma Modulator" (BDSM) which operates only with nonnegative signal with positive feedback and binary threshold. This important modification to the conventional delta-sigma modulator makes the high-speed (>100GHz) all-optical implementation possible. It has also the capability to modify its own sampling frequency as well as its input dynamic range. This adaptive feature helps designers to optimize the system performance under highly noisy environment and also manage the power consumption of the A/D converters.

The key feature that gives photonic crystals (PhCs) their ability to form photonic band gaps (PBGs) analogous to electronic band gaps of semiconductors is their translation symmetries. In recent years, however, it has been found that structures that possess only rotational symmetries can also have PBGs. In addition, these structures, known as Photonic Quasicrystals (PhQs), have other interesting qualities that set them apart of their translational cousins. One interesting feature is how defect states can be created in PhQs. If the rotational symmetry is disturbed, defect states analogous to defects states that are created in PhCs can be obtained. Simulation results of these defect states and other propagation properties of planar 12-fold photonic quasicrystal patterns, and its physical implementations in Silicon-On-Insulator (SOI) are presented. The main mechanisms required to make any optical multiplexing system is propagation; stop bands and add/drop ports. With the rotationally symmetry of the PhQ causing the stop bands, line defects facilitating propagation and now these specially design defect states acting as add/drop ports, a physical implementation of an OADM can be presented. Theoretical, practical and manufacturing benefits of PhQs are discussed. Simulated transmission plots are shown for various fill factors, dielectric contrast and propagation direction. It is shown that low index waveguides can be produced using the quasi-crystal photonic crystal pattern. Fabrication steps and results are shown.

The beam propagation method (BPM), in both two-dimensional and three-dimensional versions, is a widely used tool for modeling optical building blocks of photonic integrated circuits (PIC) and integrated optics devices. Such optical building blocks include bent waveguides, couplers, splitters, angled waveguides, etc. Most of the time, trial BPM runs need to be executed to properly set the grid density and propagation step in order to obtain stable and repeatable results. Often, these practice runs can consume a large quantity of valuable design time and computational resources, especially when modeling devices that require short propagation steps and a very dense computation grid. We propose a method that helps the BPM user to quickly assess a range of values for the grid density and propagation step, which enables adequate modeling without resorting to numerous BPM runs. This straightforward and highly intuitive method is based on what we have called the Overlap Quantization Error (OQE). It is also independent of the BPM algorithm used for the simulations. To illustrate the technique, several simulation results are presented for both high- and low-contrast curved waveguides.

This paper presents a mathematical model to predict the dynamic behavior of a MEMS based variable optical attenuator (VOA) and compares with ANSYS simulation values. The optical attenuation is achieved by the electrostatic actuation of a metal coated cylindrical waveguide. The waveguides are suitable for micromachined configurations. Electrostatic modeling employed in this analysis also takes into consideration the cylindrical geometry of the waveguide. The modeling utilizes Rayleigh-Ritz energy method in evaluating the fundamental and the higher natural frequencies of the system. The predicted natural frequencies of the optical system have been compared with ANSYS simulations to validate the proposed model. The variation of the dynamic performance of the system with respect to the critical design parameters, such as, applied voltage, electrode gap and length of the actuator is also presented. The results clearly indicate the applicability of the proposed method to optical attenuation.

This work presents the use of longitudinal refractive index modulation (apodization) in photosensitive glass for improved sidelobe suppression in volume holographic optical elements. We develop a numerical model for both uniform and apodized volume holograms based on rigorous coupled-wave analysis. We validate the model by comparison with a transmissive 1.55-μm uniform volume grating in photothermorefractive glass. We then apply our numerical model to calculate the spectral response of apodized gratings. The numerical results demonstrate that apodization of the refractive index modulation envelope improves spectral selectivity and reduces the firstorder side-lobe peaks by up to 33 dB. We suggest a method for creating apodization in volume holograms with approximately Gaussian spatial refractive index profile. We Also computed the group delay for a beam with finite transverse width. The results show the delay between the paths increases as the diffraction angles increase.

The self-image effects of the multi-port multimode couplers based on weakly guided graded Ti:LiNbO₃ optical waveguides are analyzed through the three dimensional (3D) full-vector beam propagation method (FVBPM). By applying the perfect matching layer (PML) boundary condition, we can efficiently calculate the various performances such as insertion loss, crosstalk, and power splitting ratio of the multimode couplers. Through this model, the effect of the design and fabrication parameters such as the thickness and width of the Ti-strip and diffusion time and temperature on performances of the multimode couplers can be easily investigated. By comparing with traditional MMI couplers based on step-index optical waveguides, some salient features are observed and the related design rules should be revised. Finally, some related issues such as the bending loss and high order modes are discussed.

In this paper, we present a series based solution of plane wave scattering from multiple Isosceles Right Triangle (IRT) grooves in perfect conducting plane. Scattered field in the upper half plane is expressed in a Fourier integral form. Fields in the IRT grooves are formulated as a summation of modal fields similar to the modal fields in IRT waveguide. A summation of a complete set of plane wave in the IRT groove is used to find a closed form of the fields in the IRT grooves. Matching the fields at the interface between the IRT grooves and the upper half plane provide a Fourier expansion summation form of the angular spectrum of the scattered field. The method is rigorous, robust, and provides an analytical form of the scattered field. Simulation results for far and near-fields will be shown for general oblique incident angle with various groove dimensions. Effects of the number of grooves on the scattered field are studied. The effect of ratio between the groove opening and the period between grooves is studied.

Cadmium Telluride (CdTe) is a well known photonic material in the fields of infrared imaging and solar cells. Its nonlinear optical properties also make it a promising candidate for novel telecom applications that would utilize its high Kerr coefficient to produce advanced logical devices such as switches, routers and wavelength converters. The large photorefractive effect observed in CdTe also makes possible high-speed devices suitable for optical data processing. In order to advance such photorefractive waveguide applications, we have deposited CdTe films on silicon substrates with a native oxide layer using the pulsed laser deposition technique (PLD). Silicon was chosen as the substrate material as it is suitable for the monolithic integration of logical devices. Maintaining an oxide layer was deemed necessary as a high refractive index mismatch is desirable for high-index contrast waveguide based applications and such an index mismatch could be provided by a reasonably thick layer of SiO₂. Films exhibiting some structural deficiencies, but with high optical quality were deposited through the optimization of the growth parameters. X-ray diffraction data indicates that the films are [111] oriented with rocking curves of substantial width. Atomic force microscopy images confirm that the films have a smooth surface morphology as was suggested by their mirror-like appearance. Using the optimum growth conditions, CdTe films doped with germanium were also deposited as this dopant introduces deep donor levels that enhance the photorefractive effect. A comparison of the optical properties obtained from the doped and undoped films indicate that impurities can have a marked effect on the index of refraction and extinction coefficient. Such alterations to the optical constants must be considered in the design of waveguide structures.

A relatively cheap and compact deposition system, providing excellent high temperature superconducting films parameters, has been developed and studied. Y-Ba-Cu-O thin films were deposited routinely by the YAG double-laser system with the special optical beam and plasma plume shaping systems. Buffer CeO₂ layers were deposited by the laser system as well. The properties of the plasma plume generated by the laser beam with a specific intensity distribution were investigated. A correlation between the superconducting film properties and the thermal schedule of the deposition process, the geometry of the set-up, the laser beam intensity distribution, the plasma plume shape and oxygen pressure has been found. Under optimal conditions Al₂O₃ - CeO₂ - YBa₂Cu₃O_7-δ - structures have provided the maximum critical current density of close to 10⁷ A×cm^-2 at 77 K in zero magnetic field. The films were used for the ultra high-frequency filters for the telecommunication devices.

In this study, we report for the first time the epitaxial growth of CBN thin films on Magnesium Oxide (MgO) substrates for optical device applications. A high deposition temperature (greater than or equal to to 800 ^oC) is required to obtain the epitaxial growth of CBN films. A parametric study is proposed in order to elaborate CBN thin films with a crystal structure as close as possible to that of a CBN bulk single crystal and with good optical characteristics. In particular, a low oxygen pressure (1 mTorr) allows synthesizing high-quality CBN thin films with an out-off plane lattice parameter comparable to the one of CBN bulk material at low surface roughness. The optical characterization of the high-quality CBN thin films reveals a high optical transmission (greater than or equal to 85 %) and a refractive index equal to 2.22 at 1.55 μm for certain deposition conditions. These optical properties clearly indicate the potential of CBN thin films for waveguide applications. This work presents a significant first step toward the integration and the potential use of CBN films for optical device applications.

UV transparent coatings have been used on optical fibres for direct writing of Bragg gratings without stripping-off the coating. In this paper we demonstrate a phase-mask as part of the coating of an optical fibre. We have fabricated these UV transparent, polymer phase-masks by moulding them on a holographic glass phase-mask in contact with optical fibres. These translation insensitive phase-masks have been used to fabricate fibre Bragg gratings by direct exposure to 266 nm UV radiation. Because of the nature of the polymer, the period of the phase-mask can be tuned by simple mechanical elongation, therefore changing the inscribed Bragg wavelength of the grating. We report the writing of multiple Bragg gratings in a fibre using the same polymeric phase mask. We have also used these phase-masks as stand-alone stretch tunable devices for use in a UV interferometer. These phase-masks are versatile and very cheap to replicate as they require a single phase mask. With a tunability of nearly a 1000 nm, we can write a grating with a Bragg wavelength anywhere in the full C and L telecommunication band, using a single phase mask. We believe that this technique will open new possibilities in sensing and Bragg grating fabrication of chirped and other novel filters by local deformation of the phase-mask. The concept of "dark gratings" will allow gratings to be fabricated without stripping the fibre at some time in the future, as and when required.

The development of monolithically integrated optoelectronics in silicon has been hindered to date primarily because of silicon's inefficient optical emission properties. Recently, however, nanostructured systems exploiting quantum confinement effects have shown the potential to circumvent this problem. In this study, silicon-rich silicon oxide (SiO_x, x<2) thin films doped with erbium have been deposited on silicon substrates by electron cyclotron resonance plasma enhanced chemical vapour deposition (ECR-PECVD). The formation of silicon nanoclusters along with optically active erbium ion complexes during high temperature annealing results in strong erbium photoluminescence near a wavelength of 1.54 μm. A portion of the deposition parameter space for the ECR-PECVD system has been mapped in an attempt to optimize the films for this luminescence. The resulting films ranged in composition from 0% to 22% excess silicon and 0.45% to 3.7% erbium, as determined by Rutherford Backscattering Spectroscopy. The effects of annealing were investigated between 600 ^oC and 1000 ^oC under flowing nitrogen gas. The 1.54μm emission was found to be enhanced by the presence of excess silicon, reaching a maximum at ~5-8 atomic % excess and an 800 ^oC anneal. This result strongly suggests the sensitization of infrared, erbium luminescence by silicon nanoclusters. The films exhibited an additional blueviolet light emission which has also been attributed to the erbium dopant. The visible and infrared luminescence signals were found to occur in inverse proportion to each other with the visible signal decreasing as the amount of silicon excess increases.

Earlier this year, a new breed of Spectrometers based on Micro-Electro-Mechanical-System (MEMS) engines has been introduced to the commercial market. The use of these engines combined with transform mathematics, produces powerful spectrometers at unprecedented low cost in various spectral regions.

In micro domain, thermal actuators are favored because it provides higher force and deflection than others. This paper presents a new type of micro thermal actuator that provides rotary motion of the circular disc shaped cold arm, which can be used in various optical applications, such as, switching, attenuation, diffraction, etc. The device has been fabricated in MUMPS technology. In this new design, the hot arms are arranged with the cold disc in such a way that thermal expansion of the hot arms due to Joule heating, will make the cold disc to rotate and the rotation is unidirectional on loading. The dominant heat transfer modes in the operating temperature zone are through the anchor and the air between the structure and the substrate because of the very low gap provided by MUMPS. A mathematical model was used for predicting steady state temperature profile along the actuator length and rotational behavior of the cold disc under different applied voltages. A 3-D coupled field finite element analysis (FEM) for the device is also presented. A FEM analysis was done by defining an air volume around the structure and substrate below the structure. Results obtained from the mathematical model, was compared with that of the finite element analysis. The presented results confirm the applicability of this novel rotary type thermal actuator for many optical MEMS applications.

Micro-electro-mechanical systems (MEMS) by definition are coupled electrical and mechanical microsystems. Additionally, microfabrication tolerances, device geometry and thermal effects, for example, will further cloud the performance characteristics. Hence, the consolidation of these individual parameters into a single output based upon "forward-step" modeling will allow for a complete performance characterization in a manner where changes to the static and dynamic outputs are monitored in a step wise fashion through the addition of the individual parameters separately. This deterministic approach aims to synthesize the "parameter-matrix" under which the microsystem is constrained, both by device design and by the eventual operating conditions. The theoretical modeling of the synthesized parameters into an output determinant would be a valuable design tool especially when targeting specific performance characteristics at the design stage of the microsystem that are tied to both the device design and operating conditions. This paper presents a method for microsystem performance modeling based on the solution of a parameter-matrix into a deterministically synthesized output response. The mathematical modeling is based upon the Rayleigh-Ritz energy method using boundary characteristic orthogonal polynomials. The synthesized output models the static and dynamic response of the step-forward addition of individual microsystem parameters, which when they have been evaluated can be used to specify design criteria under a given set of operating conditions. This analysis method will not only allow the designers of microsystems to determine the influence of intrinsic and extrinsic limitations and conditions, but also to establish viable MEMS platforms based on predetermined output performance characteristics.

The static and dynamic characteristics of micro-electro-mechanical-systems (MEMS) can be influenced through the application of an electrostatic field or thermal gradient. Both of these mechanisms will affect the performance of the MEMS device significantly. The thermal effects manifest themselves by varying the structural characteristics, Young's modulus of elasticity of the waveguide structure, and the material properties. These types of influences will affect the mechanical integrity through an increase in the flexibility leading to variations in the static deflections and also to the dynamic frequency eigenvalues, and changes to the device geometry can lead to faulty measurements where capacitive sensing is employed. Hence, thermal variations in the operating environment can result in unwanted thermal noise and degradation of signal integrity. Electrostatic fields or forces can be used to correct for thermal influences, for example, or as stand-alon microsystem performance tuners. The corrector characteristics can be achieved by the integration of a suspended electrode over the waveguide, for example where the induced electrostatic stiffness is aligned with the mechanical stiffness of the waveguide and are opposite in direction to the thermally induced "softening". The "stand-alone" characteristics of an applied electrostatic field can be used to selectively deflect the waveguide through an applied bias voltage and hence the static and dynamic performance can be trimmed or tuned by the application of an electrostatic field. This paper presents an experimental and theoretical investigation into coupled thermo-electrical influences on a microcantilever structure. These combined influences are typical of the operating characteristics and environments of microsystems currently in use.

In this paper we present a variable optical attenuator (VOA) that is based on microbending of a silicon-on-insulator (SOI) rib waveguide. Optical attenuation is achieved by etching away the underlying SiO₂ layer in a section of the waveguide and using electrostatic deflection to introduce vertical bending. When a single-mode rib waveguide is bent, the light traveling through it will undergo mode conversion. The amount of energy transferred to lossy modes depends on the curvature of the bending section. This mechanism is studied with the help of beam propagation method (BPM) simulations. In order to achieve a substantial amount of attenuation by bending, voltages in excess of the pull-in threshold are used, bringing a portion of the waveguide into contact with the underlying silicon substrate. An electrostatic zipping action determines the bending radii and the length of the contact region. The equation for the relationship between the bending radius of the waveguide and the controlling voltage is established through the energy method, and is numerically solved. FEM modeling is also performed to validate the result from the energy model. The device is fabricated by conventional silicon processing steps, plus steps to solve the stiction problem. The test setup for the device consists of a home-made interference microscope to monitor the vertical movement of the waveguide under test and align the input fiber to the waveguide, and other instruments to monitor the output from the device and perform the attenuation measurement. The experimental data agree well with both the BPM simulations and the calculations for the zipping actuation. The tested devices show that we can achieve 14dB attenuation over a 1mm span of a bent waveguide.

The effect of thermal tuning on the optical properties of an SOI based suspended waveguide is analysed. This analysis is based on the model that a fixed-fixed suspended beam, which forms the optical waveguide, will buckle when thermal expansion causes an axial stress that exceeds the critical buckling pressure of the beam. The analysis of the waveguide response will be broken up into the pre- and post-buckle stages of thermal actuation. Each stage of actuation will have a separate relationship for the shift in optical response as a function of temperature, which will include a combination of the thermo-optic, photo-elastic, and thermo-elastic effects. Given a corrugated, or "Bragg grating" version of the waveguide, it will be shown that thermally tuning the Bragg wavelength involves a change in index via a change in temperature and stress, and a change in grating pitch via a change in temperature. Particular attention will be paid to the evolving stress field over the length of the waveguide and its relationship to the stress-optic effect. It will also be shown that the pre- and post-buckle temperatures are path-dependant. Finally, examples of device implementation will be explored.

Light absorption spectrum measurements and the light intensity dependence of the light absorption spectrum of a fiber with a very thin gold film at the glass core glass cladding boundary are presented. The thickness of the gold film is less than the scattering length of electrons in this metal. The absorption spectrum appears to be strongly light intensity dependent. We also observed the mode structure of light propagating through the gold film. Our fabrication process can produce large area very thin metal films that are very difficult to produce by other methods.

Quantum operator algebra related to Jaynes-Cummings model is developed to design multi-reflector resonant bandpass filters for the first time. Transmittance and reflectance spectra of these filters are givens with analytic expressions. The results are found in agreement with those based on the existing filter design method. By selecting parameters such as r and N, designed filters can achieve a target spectrum profile with flat-top, large bandwidth, and minor ripples.

In this paper, we reported the crystal growth and optical properties of Yb³⁺ and Er³⁺ codoped yttrium scandium gallium garnet (YSGG) crystal by the conventional CZ method. The growth conditions and XRD patterns are given out.. Excitation and luminescence spectra were also recorded at room temperature with a Model F111Ai spectrofluorometer using a 980nm LD laser as an excitation source. The results were discussed in detail.

Optical technology holds considerable promise to improve early detection, diagnosis and risk assessment of breast cancer. Unlike current clinical risk assessment tools such as the Gail model, the most widely accepted risk assessment tool, optical risk assessment technology can be applied to the entire female population of all ages. This study is investigating the use of optical reflectance spectroscopy (ORS) as a possible breast tissue development monitoring tool for adolescent girls. Changes in breast development due to proliferation of mammary gland and the surrounding stroma are reflected in changes in breast tissue density and composition which can be interrogated optically. Modifications of development influenced by micronutrients and hormonal status from exposures (e.g. toxins), lifestyle and diet effects, may ultimately be tracked. Preliminary data suggests that ORS has the ability to detect differences in bulk tissue properties in the developing breast of adolescent girls when compared to developmental stages assessed by Tanner, regional variation within breast tissue structure and asymmetries between left and right breast size and shape. Spectral comparison of unilateral breast development permits adjusting the optode separation as function of developmental breast size to minimize optical sampling of pectoral muscle.

Multimode interference (MMI) couplers are important integrated optical components for the optical signal processing and routing. The realization of these components by ion exchange on glass substrates is particularly attractive for low cost integration. The design and analysis of MMI devices have generally been based on the self imaging principle in step-index waveguides, whereas waveguides fabricated by ion exchange on glass are practically graded-index due to the nature of the thermal diffusion of exchanged ions. In addition, the ion exchange process results in a guide with depth that depends on the mask opening (the guide width) which causes a high insertion loss at the interface between single mode and multimode sections of the MMI. To overcome these problems 3D simulation of the ion exchanged MMI structures is strongly required. In this work such 3D simulation is achieved on two levels. First the non-linear diffusion equation describing the ion exchange process is solved numerically using a finite-difference method with a modified algorithm to ensure solution stability for an extended range of nonlinearity. The resultant index distribution is used in a wide angle 3D BPM to simulate the optical field propagation in the structure. This allows accurate prediction of the structure performance under different fabrication and excitation conditions. Based on this simulation technique, 3 dB MMI splitter design with tapered access guides is optimized by both geometrical mask design and process parameter variations. The optimization shows that both the tapering and the use of annealing process can significantly improve the performance of the devices.

A sensitivity enhanced long-period grating (LPG) refractive index sensor is proposed and studied by using the LP model. In the simulation, the cladding layer of the LPG is assumed to be partially removed and then deposit a sensitivity enhancement layer (SEL) with a higher refractive index. The effects of the thickness of the original cladding material, and the thickness and refractive index of the SEL layer on the LPG transmission spectrum notch wavelength shift as a function of the ambient refractive index change are reported. The LPG sensor performance depends on the phase match condition of the core mode and cladding mode coupling in the LPG structure and the dependence of the effective index of the cladding mode on the thickness of the original cladding material, and the thickness and refractive index of the SEL layer. The structure modified cladding layer moves the work point of the long period grating to the cladding mode reorganization zone, where the cladding mode effective refractive index changes rapidly upon the LPG waveguide parameter. Proper selection of parameters of the cladding layer can be applied to enhance the modulation of the effective index of the cladding mode by the ambient refractive index through the evanescent field and thus construct sensitivity enhanced ambient refractive index sensors.

We present a fast and efficient numerical model for Yb³⁺-doped fiber lasers based on shooting method. The algorithm is based on the assumption of a starting value for the slope efficiency and the evaluation of the pump power threshold. The starting value of slope efficiency is related to initial conditions through the boundary conditions, and it is subsequently optimized by iteration. The method ensures a fast and efficient convergence of the solution of the coupled first-order differential equations that describes the evolution of pump and signal powers in a Yb³⁺-doped fiber laser. The results of the numerical solution are compared with experimental and published data giving a good agreement.

UV light from an Excimer laser operating at 248 nm with fluence of 800 mJ/cm² per pulse at 30 Hz repetition rate is used for writing process of a fiber Bragg Grating (FBG). Multiple FBGs with different center wavelength and reflectivity are produced by varying strain and exposure time on a single strand of a SMF where 1552.75 nm is the center wavelength of the unstretched FBG. The in-situ measurement of the reflectivity, bandwidth and center wavelength of each Bragg grating will be described. We will also present an RF technique for measuring the separation between two adjacent FBGs.

We develop half-wave retarder based of deformed-helical ferroelectric (DHF) liquid crystal cell. Birefringence and thickness of DHF cell is optimized to obtain ninety degrees rotation of the light polarization. An optical scheme of crossed-switch between two inputs and two outputs channels using such DHF cell is discussed. The switching time, less than 10 microseconds at controlling voltage of 20V can be provided, which is temperature independent over the broad temperature range.

Ultrafast-laser micromachining has promise as an approach to trimming and 'healing' small laser-produced damage sites in laser-system optics--a common experience in state-of-the-art high-power laser systems. More-conventional approaches currently include mechanical micromachining, chemical modification, and treatment using cw and long-pulse lasers. Laser-optics materials of interest include fused silica, multilayer dielectric stacks for anti-reflection coatings or high-reflectivity mirrors, and inorganic crystals such as KD*P, used for Pockels cells and frequency-doubling. We report on novel efforts using ultrafast-laser pulsetrain-burst processing (microsecond bursts at 133 MHz) to mitigate damage in fused silica, dielectric coatings, and KD*P crystals. We have established the characteristics of pulsetrain-burst micromachining in fused silica, multilayer mirrors, and KD*P, and determined the etch rates and morphology under different conditions of fluence-delivery. From all of these, we have begun to identify new means to optimize the laser-repair of optics defects and damage.

New methods, implemented in CMOS, mitigate the effects of analog degradations on 10 Gb/s optical signals. Experimental results, across 320, 1280, and 5280 km of standard single-mode fiber, without any optical compensators, show complete compensation for the dispersion and substantial elimination of the SPM.

The regenerative properties of using a clock-modulated pump for higher-order four-wave mixing in a highly nonlinear fiber are demonstrated through simulations and experimental measurements, at 10 Gb/s, for 3R optical regeneration and NRZ-to-RZ format conversion. The use of a clock-modulated pump for higher-order four-wave mixing enables retiming and enhances the reshaping properties of the signal, thereby improving the regeneration of the signal, relative to a continuous-wave pump.

The performance of an all-optical regenerator utilizing self-phase modulation in a highly nonlinear fiber and offset optical filtering is assessed using computer simulation. By varying the bandwidth and offset of the optical filter, the Q-factor performance of the regenerator is near-optimized for systems impaired by ASE noise and systems impaired by both residual dispersion and ASE noise. Generally, the near-optimum bandwidth and offset of the optical filter differs for these two types of systems. It is found that the selection of the bandwidth and offset is more stringent for systems with ASE noise only. For systems with residual dispersion and ASE noise, the selection of the filter bandwidth and offset depends on the amount of residual dispersion with quite different trends being observed. The regenerator is more effective when the residual dispersion is negative.

Multimode fiber (MMF) has found applications in high-speed computer interconnect, local area networks (LAN), and storage area networks (SAN) due to its ease of handling and high performance over short span. However, modal dispersion limits its bandwidth-distance product (BDP) to about 2 Gb/s-km. This limit has been extended by recent new generation of optimized MMF to 28 Gb/s-km, but there is evidence that a substantial portion of installed MMF have imperfect refractive index (RI) profiles due to defects during the manufacturing process, and the BDP might be at best no more than 500 Mbps-km. Different strategies have been proposed to address this issue by employing offset launch, multi-level subcarrier modulation, and mode spatial control. However, our studies have shown that end-to-end system performance of installed MMF can be highly dependent on input launch polarization. In this report, we investigate, for the first time to our knowledge, the relationship between RI profile defect, input launch condition, and transmission performance in commercial-grade MMF, both 50 μm and 62.5 μm. To this end, a number of techniques have been deployed. Two-dimensional (2D) MMF RI profile is obtained by a micro-reflectivity technique with a spatial resolution of ~400 nm. MMF transmission characteristics are interrogated using interferometric techniques. Data at 40 Gb/s are transmitted over the same MMF sample at different launch conditions, and the system performance is evaluated by bit-error rate measurements. These results are then analyzed to provide insights to correlate fiber RI profile defects and high-speed data transmission performance for installed commercial-grade MMF for optical access networks.

The data-entropy quality-budget developed by the authors is used as an alternative to the conventional power budget. The traditional power budget approach is not capable of providing a full analysis of a system with different noise types and specifically providing a measure of signal quality. The quality-budget addressed this issue by applying its dimensionless 'bit measure' to integrate the analysis of all types of losses. A data-entropy visualisation is produced for each set of points in a reference and test signal. This data-entropy signal is a measure of signal disorder and reflects the power loss and types of signal degradation experienced by the test signal. To analyse the differences between two signals an algorithm known as phase-coherent data-scatter (PCDS) is used to assess levels of attenuation, dispersion, jitter, etc. Practical analysis of telecommunications signals using the new multiple-centroid (MC) PCDS is presented here for the first time. MC-PCDS is then used to analyse differences between sets of data-entropy signals and digital signals. The theory behind MC data-scatter is discussed and its advantages for the quantification of signal degradations are assessed. Finally, a brief consideration is given to the use of pattern recognition algorithms to measure optical signal degrading factors.

Small form factor pluggable (SFPs) transceivers have been used to demonstrate that legacy optical networks operating in the metro-core-edge region of the network can be upgraded cost effectively through optical-electrical-optical (OEO) wavelength translation (WT). The WT technique involves detection of the original communication signal, re-timing and retransmission at a new laser wavelength and linewidth. The change is a deliberate one and gives rise to better optical transmission characteristics, resulting in longer system reach. In this paper, we present experimental results that demonstrate how WT can be used to upgrade existing optical fiber communication systems through improvements in system dispersion by using dense wavelength division (DWDM) SFPs and single-channel dispersion-compensation modules (DCM).

With the advent of new regulations governing the protection and recovery of sensitive business data, including the Sarbanes-Oxley Act, there has been a renewed interest in business continuity and disaster recovery applications for metropolitan area networks. Specifically, there has been a need for more efficient bandwidth utilization and lower cost per channel to facilitate mirroring of multi-terabit data bases. These applications have further blurred the boundary between metropolitan and wide area networks, with synchronous disaster recovery applications running up to 100 km and asynchronous solutions extending to 300 km or more. In this paper, we discuss recent enhancements in the Nortel Optical Metro 5200 Dense Wavelength Division Multiplexing (DWDM) platform, including features recently qualified for data communication applications such as Metro Mirror, Global Mirror, and Geographically Distributed Parallel Sysplex (GDPS). Using a 10 Gigabit/second (Gbit/s) backbone, this solution transports significantly more Fibre Channel protocol traffic with up to five times greater hardware density in the same physical package. This is also among the first platforms to utilize forward error correction (FEC) on the aggregate signals to improve bit error rate (BER) performance beyond industry standards. When combined with encapsulation into wide area network protocols, the use of FEC can compensate for impairments in BER across a service provider infrastructure without impacting application level performance. Design and implementation of these features will be discussed, including results from experimental test beds which validate these solutions for a number of applications. Future extensions of this environment will also be considered, including ways to provide configurable bandwidth on demand, mitigate Fibre Channel buffer credit management issues, and support for other GDPS protocols.

The RISQ (Réseau d'Informations Scientifiques du Québec) research and education network has been a trail-blazer in the development of privately owned fiber networks, by implementing an optical fiber owner model. This model allows RISQ to have a gracefully upgradable network that is at the forefront of the large bandwidth optical communications technology. It also allows RISQ to offer a parallel network that can be used as a test equipment for the R&D community, in universities, research institutions or the industry; a real in-the-field fiber optic network test bench. RISQ is in a privileged situation to influence optical communications R&D work. It is aware of the real needs of the telecommunications industry. It is also aware of the technological needs of the next-generation networks. RISQ suggests that telecommunications R&D should be focused on increasing network reliability and decrease network operations and capital expanses, rather than on increasing their capacity. A key to decreasing expanses would be to avoid SONET network management on long-haul trunks, without affecting the transmission quality of service. IP over optics should thus be reinstalled as a priority in the telecom R&D world. RISQ thinks optical 3R might be the solution that will allow IP over optics. As for the next-generation networks: flexibility, reconfigurability, in order to offer lightpaths of adjustable bandwidth to the user; while wasting a minimum of the valuable network bandwidth, is where we believe efforts should be concentrated on.

An analytic model is reported to evaluate the electric output signals and their variance (Eye Diagram) for multi channel high-speed DPSK fiber optical system in presence of polarization mode dispersion (PMD), polarization dependent loss (PDL) and chromatic dispersion (CD). It is also found that even under linear cross talks, the balanced receiver output show a strong asymmetric phenomenon if adjacent channels are non synchronous.

In this paper, polarization properties of spun fibers with millimeter spin periods and certain intrinsic birefringence are investigated both theoretically and experimentally. Polarization evolution and modal coupling in spun fibers are studied by using the Jones vector method, and fiber polarization-mode dispersion is examined by using the Jones matrix eigenanalysis method. Theoretical results are in good agreement with the measured results. Some potential applications of these spun fibers for optical communications and fiber sensors as well as the related requirements are discussed.

A polarization dependent loss (PDL) vector equation of motion in Stokes space was derived in a system interacting with polarization mode dispersion (PMD). A new PDL measurement method based on the PDL vector equation was proposed and validated by numerical simulation and experiment.

We have developed a tunable multi-wavelength semiconductor fiber laser (SFL) for chromatic dispersion measurements of optical fiber based on the time-of flight method. The SFL incorporates a programmable high-birefringence fiber loop mirror to select the separation of the lasing wavelengths between 3.2nm and 1.6nm. The SFL emits 5 wavelengths with an average power of 11.96dBm per wavelength and 11 wavelengths with an average power of 18.35dBm per wavelength, for separations of 3.2nm and 1.6nm respectively, all within the C-band. The linewidth of each oscillating wavelength resides in the 0.16nm - 0.28nm range, the signal-to-noise ratio varies between 33.5dB and 39.2dB, and the uniformity of the output power is within 3.2dB. Stability measurements for each lasing peak show a wavelength deviation of +/-0.09nm/hour and a power variation of +/-0.90dB/hour. Results from time-of-flight measurements are compared with standard phase-shift techniques and the differences analyzed. The percent error between the two methods is better than -0.73 to 1.13% for measurements on various standard optical fiber lengths. The time-of-flight method is easier and faster to use for the characterization of sufficiently dispersive media such as deployed fiber spools. Our tunable laser provides a simple low cost solution for such measurement applications. The tunable nature of our SFL source also provides the following advantages for chromatic dispersion measurements: (1) greater precision can be obtained since two independent measurements (one at each wavelength separation) can be performed using a single optical source and (2) there is increased flexibility since the wavelength spacing can be tailored for a specific situation; shorter lengths benefiting from larger wavelength separations.

Measurement of optical power loss in a fiber splice presents several challenges for manufacturers of optical modules. Requirements include measurement sensitivity of <0.05 ±0.005 dB for ultra low loss fusion splices between similar single mode fibers, and acceptable repeatability and reproducibility (R&R). These issues and several important gaps in the current loss measurement standards are addressed in a new draft standard, recently submitted to the Telecommunications Industry Association (TIA). The new standard describes measurement methods appropriate for various applications, including splices made with similar and dissimilar fiber types, and attenuating fiber at the test wavelength. This paper discusses the loss measurement results and gage R&R comparisons for the different methods included in the new standard, the practical aspects of making measurements and the question of directionality for dissimilar fiber splices. The experimental work was undertaken by the iNEMI (International Electronics Manufacturing Initiative) Fiber Optic Splice Improvement Project. The new loss measurement standard is a collaboration between iNEMI and the TIA.

A technique is presented for measuring the optical phase transfer function of optoelectronic devices for stimulus frequencies from 100 MHz up to the modulation bandwidth of the device. Using high-resolution optical spectra and a novel instrument setup that makes use of RF signal processing to obtain the stimulus signals, the change in phase of an optical signal is obtained as a function of a time-varying electrical stimulus for electrical-to-optical devices or an optical stimulus for optical-to-optical devices. In the technique, the modulation frequency of the stimulus can be varied over a wide range (e.g., 100 MHz to 10 GHz for a 10 Gb/s device). Thus the proposed technique complements low-frequency and stepped measurements of the optical phase transfer function. The technique is demonstrated by considering the change in phase of the output signal from an electroabsorption modulator as a function of the applied voltage.

In the paper, we present a novel method to measure modal power distribution (MPD) in few-mode and multimode fibers using embedded tilted fiber Bragg gratings (TFBG). The TFBG can couple portion of the guided modes into the corresponding radiation modes, whose powers can be selectively measured through a spatial filter. For few mode fibers, the power coupling coefficients between guided modes and their corresponding radiation modes are obtained by solving a set of linear coupling equations acquired under different launching conditions at the fiber input. For multimode fibers, the power coupling coefficients can be measured separately under single mode-group excitation condition. Then, the powers of guided modes in a few-mode or multimode fiber under any excitation condition can be obtained by simply measuring the powers of radiation modes and calculated using the solved coupling coefficients. The proposed method is successfully demonstrated.

Light can interact with periodic microstructures, also known as photonic crystals, in many ways. In this talk we will consider the use of photonic crystals to modify the wavefront and wavevector direction of waves propagating in photonic crystal slab waveguides and will compare these effects with those at arise from modification of the Poynting vector. We will discuss the underlying principles and will consider applications of these devices, in particular for wavelength demultiplexing. We will illustrate our investigation by considering the design of 1-D and 2-D photonic crystal superprism demultiplexers, and show that compact (<4 mm chip size) and high resolution (100 GHz) multichannel devices can be obtained.

Until recently, reconfigurable optical add/drop multiplexer (ROADM) systems did not exist, their components were unselected, and their market was unclear. Today, every major system vendor has a ROADM offering, and a large number of component vendors have announced ROADM products based on a variety of technologies, some more mature than others. We review the different optical component technologies that have been developed for use in ROADM subsystems, and describe their principles of operation, designs, advantages, and challenges. The technology platforms that we cover include MEMS, liquid crystals (liquid crystal devices (LCD) and liquid crystal on silicon (LCoS) technologies), and monolithic and hybrid planar lightwave circuits (PLC) based on silica on silicon and polymer on silicon platforms. For each technology, we describe the corresponding ROADM subsystem architectures in terms of functionality, features, size, cost, and maturity.

Optical add/drop multiplexers (OADMs) have emerged as the key enabling components for building long-haul and metro-area networks. The wide-spread deployment of OADMs in the access market will depend on the availability of cost-effective integrated solutions. We have successfully fabricated a fully-integrated OADM based on planar reflective gratings. The device uses a combination of two grating elements arranged in a subtractive dispersion configuration. The first grating demultiplexes a 300-nm-wide band and drops optical channels at 1490 nm and 1550 nm, commonly used by service providers to send information to the end user. The second grating completely counter-balances the dispersion properties of the first grating and ultimately yields zero dispersion in the output waveguide. Such a configuration allows the transmission of optical signals though the OADM in an ultra-wide band spanning 1250 to 1410 nm. This ultra-wide 'through' band is a critical step allowing the use of low-cost lasers, without temperature stabilization, for sending data to a service provider. The OADM was manufactured using an industry standard silica-on-silicon process which was augmented with grating facet formation and metallization. In spite of using low refractive index contrast waveguides (0.82%), the device had a remarkably low footprint of only 0.25 square centimeters. Applications of the OADM in access market networks is discussed.

Spurred by the growth of the internet, Optical Telecommunications bandwidth, experienced unprecedented growth during late 1990's. During this time of great economic expansion, the creation of new enterprises was vast and the expansion of established component, system and services companies was breathtaking. Unfortunately, this positive economic state was short-lived. This period was followed in 2001-2004 by one of the most significant market crashes in history. During those 10 years of economic growth, about $20B in venture capital was invested in the optical telecom industry, most of this investment was lost in recent years. Many start-up industries which experienced unprecedented growth at the end of the 20^th century were lost at the start of the 21^st. (1) During this time many, innovative technologies were born and buried. However, many new capabilities emerged from this period of unrest; one such example is the advent of Optical MEMS (MOEMS). Many academics and corporate laboratories pursued the development of MOEMS during the economic boom and, in the author's view; MOEMS surfaced as a powerful and versatile tool set that has proved invaluable and in the last few years during economic downturn, stood the test of time. In the Telecommunications industry, for optical switching and wavelength management applications MOEMS has proven to be the technology of choice. (2) Variable Optical Attenuators (VOA), Wavelength Blockers (WB), Dynamic Gain Equalizers (DGE), and most recently Wavelength Selective Switches (WSS) are being used in the numerous recent network deployments. Moreover, agile networks of the future will have MOEMS at every node. This presentation will provide an overview of the history of MOEMS in Telecommunications, discuss its byproducts and offer a window into the future of the technology.

Compact optical sources that generate picosecond pulses at multiple wavelengths are of interest for numerous applications in optical instrumentation, fiber optic sensing, and optical communications. In recent years, numerous methods have been demonstrated to obtain multi-wavelength, mode-locked (ML) operation from erbium-doped fiber lasers (EDFLs) and semiconductor fiber ring lasers (SFRLs). In contrast to EDFLs, the use of semiconductor optical amplifiers (SOAs) allows for stable, multi-wavelength emission at room temperature with narrow wavelength separation since they are not constrained by the EDF homogenous broadened linewidth, and for operation over a wide wavelength band. To increase the functionality for some applications, it is also important to be able to tune the output wavelengths of the optical pulse source. In this paper, we provide an overview of our on-going work on developing tunable multi-wavelength, ML-SFRLs. In terms of achieving multi-wavelength operation, we have used multi-wavelength filters based on a high-birefringence Sagnac loop and superimposed fiber Bragg gratings. In terms of short pulse generation, we have explored two different methods for mode-locking: the use of an intra-cavity electro-optic modulator and the injection of an external optical control signal to modulate the gain of the SOA via cross-gain modulation. Finally, in terms of wavelength tunable operation, we have exploited dispersion tuning, i.e. the use of a dispersive cavity and changing the modulation frequency of the mode-locking element. We present and discuss our results for two different ML-SFRL configurations.

This report presents an investigation of composite fiber Raman amplifiers, i.e. a distributed fiber Raman amplifier followed by a discrete fiber Raman amplifier, both with incoherent pumping, compared to conventional coherent pumping. It is shown that a flatter gain and optical signal-to-noise ratio (OSNR) over 100-nm bandwidth can be achieved by using two incoherent counter-pumps, compared to using six coherent counter-pumps. Moreover, it is also found that further flatter gain and flatter OSNR over 100-nm bandwidth can be obtained simultaneously in composite fiber Raman amplifiers with bi-directional pumping. The flatness of both gain and OSNR with a ripple of 1 dB is predicted by using one incoherent co-pump and one incoherent counter-pump.

Two important aspects of optical networking are to ensure transmission integrity when connecting optically transparent channels in real time for protection, and the handling of dynamic demand, especially when survivability considerations are added. The prevalent approach for providing protected dynamic lightpath services is signaling intensive, database dependent, and requires on-the-fly cross-connection of standby lightwave channels to form protection paths. With two inter-related new concepts, we address concerns about on-the-fly concatenation of transparent optical channels to form protection paths and improve the scalability, and reduce database dependence and signalling intensity associated with current methods. An extension of the p-cycles concept to end-to-end path protection, in conjunction with the concept of a protected working capacity envelope for dynamic provisioning (which has no per-connection signaling for backup path establishment) yields a highly efficient network architecture in which protection paths are fully pre-cross-connected prior to failure and only the two lightpath end-nodes need to detect path failure and switch the lightpath (or its traffic) into a corresponding pre-planned path-protecting p-cycle.

Since the deployment of optical communications networks, carriers have used many mechanisms to protect fiber in the core, because of the large volume of traffic carried in the core, and the affect of a failure on the entire user community. Access links, connecting end users to the core network were not protected because they affected a small number of customers, carried relatively low bandwidth traffic and were less critical. As enterprises use ever growing bandwidth and become increasingly dependent on their network connectivity, they demand protected services. Since fiber breaks can occur anywhere in the network, including the so-called "last mile" carriers are looking today for protection systems for the access and edge network links. This paper focuses on the emerging solutions for this application, detailing the key criteria for selecting fiber protection switches: optical performance and cost effectiveness (important because the cost must be amortized over a small number of customers). It discusses solutions to the various access network topologies and common protection schemes.

This paper presents a burst aggregation method for an Agile All-Photonic Network (AAPN) operating under an asynchronous burst switched mode. The model combines both the timer-based and threshold-based approaches into a single composite burst assembly mechanism. This is evaluated semi-analytically for fixed length packets and Poisson arrivals and used as a special case to verify a more general OPNET Modeler simulation. The dependence of the blocking probability on different burst aggregation parameters is observed as well. The same procedure is extended to 'encapsulate' (aggregate) variable packet length traffic into 'envelopes' (bursts) matched to the time slots in an AAPN operating in a synchronous time-slotted mode. Results are presented for an emulation of this process using real IP network traffic from the local LAN using two encapsulation methods that differ depending upon whether 'envelope' boundaries are allowed to cross constituent packets or otherwise. Bandwidth utilization was measured for different encapsulation parameters and it is confirmed that the model with the boundaries allowed to cross packets (i.e., the model with packet segmentation) is more bandwidth-efficient even if the processing delay is slightly larger. The successful operation of the emulation system suggests as well that a simple, low-cost software implementation would be suitable to perform the burst/slot aggregation process in AAPN.

Dense Wavelength Division Multiplexing (DWDM) passive optical network represents the most promising solution for future access networks. Wavelength sharing is an important issue in DWDM based optical access networks(OAN), since it can maximize the resource utilization and in return reduce the cost of the system. In this paper we propose a neighbor-sharing dynamic wavelength assignment algorithm for a DWDM-based virtual start optical network. The wavelength assignment algorithm is realized through a centralized scheduler, which is at central office. When the scheduler receives the request, it assigns a wavelength to the corresponding premise according to the algorithm. Two level scheduling scheme of control signal is proposed. The evaluation and analysis of the performance of the wavelength assignment algorithm are conducted by using computer simulations and are presented at the end of this paper.

In optical code division multiple access (OCDMA), the optical bandwidth is accessed simultaneously by multiple users, leading to intensity noise in photodetection. In this article we focus on spectral amplitude encoded CDMA for access networks. We examine efficient system architectures, practical implementations, compare the impact of different system designs on spectral efficiency and bit error rates. We draw on both experimental and simulation results. We discuss the use of two types of optical sources, a broadband incoherent source and a multi-laser source. We examine the capacity that can be achieved for these systems and the design goals we are shooting for in our laboratory.

This paper presents the experimental demonstration of a 1 Gbps Direct Sequence Optical Code Division Multiple Access (DS-OCDMA) system using pulsed coherent source. Encoding and decoding using Prime Sequence codes are achieved by Sampled Fiber Bragg Gratings (S-FBGs). The encoders/decoders have been designed with OptiGrating software and realized with Phase Mask Process. BER measurements have been performed in the asynchronous configuration when an interferer is delayed from the desired signal. A 2 dB penalty due to Multiple Access Interference effect (MAI) has been observed in the synchronous case. S-FBG technological limitations and optical interferences due to the source coherence time have been observed and will be discussed.

Recently the wavelength division multiplexing (WDM) networks are becoming prevalent for telecommunication networks. However, even a very short disruption of service caused by network faults may lead to high data loss in such networks due to the high date rates, increased wavelength numbers and density. Therefore, the network survivability is critical and has been intensively studied, where fault detection and localization is the vital part but has received disproportional attentions. In this paper we describe and analyze an end-to-end lightpath fault detection scheme in data plane with the fault notification in control plane. The endeavor is focused on reducing the fault detection time. In this protocol, the source node of each lightpath keeps sending hello packets to the destination node exactly following the path for data traffic. The destination node generates an alarm once a certain number of consecutive hello packets are missed within a given time period. Then the network management unit collects all alarms and locates the faulty source based on the network topology, as well as sends fault notification messages via control plane to either the source node or all upstream nodes along the lightpath. The performance evaluation shows such a protocol can achieve fast fault detection, and at the same time, the overhead brought to the user data by hello packets is negligible.

The Agile All-photonic Backbone Network (AAPN) architecture has been proposed by the telecommunication industry as a potential candidate for the ultra high speed Next Generation Optical Network (NGON) architecture. AAPN network structure is composed of adaptive optical core switches and edge routers in an overlaid star physical topology. The AAPN employs fast packet switching architecture for the network traffic, and the packet scheduling is the main part of the AAPN. The objective is to forward the packets to their destination with the lowest drop rate and delay, the bandwidth allocation can be either located at the core node or the edge switch. Two types of scheduling are considered in the AAPN architecture, namely the centralized and the distributed schemes. In the centralized scheme all decisions are made at the core node while in the distributed scheme, they are made at the edge nodes. In this paper, we want to compare both scheduling schemes. We would also like to propose a promising integrated TDM architecture that combines the good attributes of both centralized and distributed scheduling schemes. We shall characterize such architecture by various measures such as delay and loss probabilities.

All-Optical networks with the DWDM technology provide huge bandwidth and are the sole approach for transporting huge network traffic. However, this bandwidth is too coarse to be used by a single user and this is why the Optical Time Division Multiplexing (O-TDM) has been deployed in the optical networks to provide finer granularity and improve bandwidth usage. In contention-based slotted-optical networks, because there is no collaboration among the ingress switches, the data slots on the same wavelength and time-slot destined to the same destination may collide. In this paper, we detail contention avoidance schemes in two software and hardware categories and show that edge switches can have an important role in reducing loss rate in optical networks by transmitting traffic to each destination with equal probability (symmetric traffic transmission) and balancing traffic load on the wavelength channels. We also show that edge switches can have an important role in the loss rate reduction issue in optical networks by reducing traffic load and using more wavelengths/fibers to carry the same traffic.

Pair control is generally adopted in telecommunication systems to save software processing time to set up both-way paths between a sender and a receiver. We have introduced new pair control architecture into a Photonic Switch Fabric (PSF) which is a key instrument of a developed Photonic Label Switch Router (PLSR). In our WAPS (Wavelength Assignment Photonic Switching System), the PLSR is to be operated as a photonic router in the optical internet system to guarantee QoS (Quality of Service) and the transmission bandwidth by providing end-to-end wavelength LSPs (Label Switched Paths). In contrast with that conventional pair controlled PSFs do not have capability to save switching elements and its control circuits, the proposed architecture enables to save them drastically.

Optical pulse shaping finds many important applications in the area of ultrafast photonics. For example, in ultrafast all-optical switching, temporal shaping of the control pulses can be used to create a wide flat-top switching window with sharp rise and fall times. Such switching windows are more jitter-tolerant than simple Gaussian windows, and can therefore achieve a lower bit-error rate. Previously reported pulse shaping methods using linearly chirped Fibre Bragg Gratings (FBGs) take advantage of the direct correspondence between the spatial distribution of the grating periods and the temporal distribution of the spectral contents of the grating impulse response (space-to-frequency-to-time mapping). However, they have two major drawbacks: first, they are only valid for high-dispersion gratings, and therefore are unsuitable for producing short pulses; second, due to the inherent impulse response assumption, the power conversion efficiency is very low since the grating bandwidth needs to be much smaller than that of the input pulse. The numerical conversion efficiency demonstrated for this method is about a few percent. We report a versatile technique for temporal pulse shaping using a simple linearly chirped FBG and an amplitude mask. Unlike previous pulse shaping methods, ours is also applicable for low-dispersion gratings with bandwidths comparable to that of the input pulse (i.e., taking into account of finite input pulse duration). The chirped grating is used to stretch the incoming pulses to the desirable temporal width, while the amplitude mask modifies the shape of the pulses. We developed a novel optimization algorithm to obtain an amplitude mask that significantly increases the conversion efficiency. Pulse shaping using linearly chirped FBGs in the low dispersion regime is simulated using two methods. For weak gratings, the direct relationship between the spatial grating profile and the grating impulse response was used in tandem with a forward-correction deconvolution algorithm to solve for the optimum amplitude mask. For strong gratings, an optimization algorithm based on the partial validity of space-time mapping, as well as the causal relationship between the reflected temporal response and the grating apodization profile was used to design the amplitude mask. We experimentally demonstrated the conversion of 1-ps transform-limited Gaussian pulses to 10-ps pulses with a target shape at a high conversion efficiency of ~20% (measured) using a 1.5-mm-long grating. The spectral width of the pulses is 3.5 nm, centered at 1.55μm. As the reflected pulse shape is controlled by the shape of the amplitude mask, our method can be easily adapted to produce any arbitrary temporal pulse shape by designing an appropriate amplitude mask. To our knowledge, this is the first demonstration of this technique of arbitrary pulse shaping using the combination of a linearly chirped fiber grating and an amplitude mask in the low-dispersion regime.

Discrete fiber Raman amplifiers (DFRAs) are a candidate technology for extending the transmission capacity of dynamic optical networks such as agile all-photonic networks. However, as with conventional doped fiber amplifiers, they suffer from power transients following the addition or removal of channels which results in system penalties due to degradation in optical signal-to-noise-ratio or nonlinear effects. Recently, several groups have investigated all-optical gain-clamping, which consists in introducing a lasing feedback signal in the amplifier, as a means for gain control and mitigating the power transients. In this paper, we theoretically analyse the dynamic response of DFRA cascades in the worst possible case of power transients. In particular, we consider a 64-channel system in which all of the channels except one (i.e. the surviving channel) are cut from and subsequently added to the first amplifier of the cascade. While previous studies have focused on the transient response of the surviving channel when all of the amplifiers in the cascade are either unclamped or gain-clamped, we consider here cascades involving combinations of unclamped and gain-clamped DFRAs. We vary the number and the position of the gain-clamped DFRAs in the cascade to determine whether a cascade in which only a few amplifiers are gain-clamped can be effective for controlling the power transients within tolerable limits. To simulate the dynamic response of the DFRAs, we have developed a new technique, based on a temporal average power analysis. In our simulations, we also take into account the location of the surviving channel and the operational regime of the amplifiers. Our results show that the location of the gain-clamped DFRAs in a mixed cascade is not important when the amplifiers are operated in small-signal regime; on the other hand, as the input signal power is increased, it becomes more preferable to place the gain-clamped amplifiers at the beginning of the cascade.

Wavelength-selective optical components used in WDM optical communication systems often exhibit a transfer function with amplitude and phase ripples. These ripples can lead to signal distortion and performance degradation. The pulse distortion induced by sinusoidal amplitude and phase response ripples is derived for Gaussian pulses, and concise results are presented for the power penalty in system performance. The analysis shows that the amplitude and phase response ripples have a similar impact on the transmitted signal, and the power penalty induced by these ripples depends on the chirp and pulse width of the transmitted signal. The combined effect of the characteristics of the transmitted signal and the ripple parameters on the system performance is discussed in detail. Numerical simulations show a good agreement with the theoretical analysis.

Optical packet switching (OPS) and optical burst switching (OBS) are regarded as next-generation transport technologies that enable more efficient and flexible utilization of the capacity of optical networks by providing sub-wavelength granularity. Optically labelled packet transmission based on orthogonal intensity-modulation/differential-phase-shift-keying (IM/DPSK) modulation format, in which the payload is intensity modulated while the label is carried by DPSK has been proposed and demonstrated. More recently, it was found that using DPSK/IM for payload/label modulation and a balanced receiver for DPSK detection is more advantageous. In these optical label encoding schemes, two optical modulators are required, one for encoding the payload and the other for the label. In this paper, we demonstrate a novel payload and label encoding technique based on a single Mach Zehnder (MZ) modulator. In this scheme, the RF port of the MZ modulator is used to encode a 10G DPSK payload while the bias port is used to impose the label information through an appropriate intensity modulation. Direct detection of the label is achieved with an inexpensive low-speed receiver while the DPSK payload is decoded by using an optical 1-bit delay interferometer before detection by either a single or a balanced detector. Experimental results show superior receiver sensitivity for both the label and the payload, which compares favourably with previous reported schemes and with the advantage of using only a single modulator. Furthermore, we show that label removal and re-insertion can be realized in the RF domain without any polarization dependence.

Driven by the world's growing need for communication bandwidth, progress is constantly being reported in building newer fibers that are capable of handling the rapid increase in traffic. However, building an optical fiber link is a major investment, one that is very expensive to replace. A major impairment that restricts the achievement of higher bit rates with standard single mode fiber is chromatic dispersion. This is particularly problematic for systems operating in the 1550 nm band, where the chromatic dispersion limit decreases rapidly in inverse proportion to the square of the bit rate. For the first time, to the best of our knowledge, this document illustrates a new optical technique to post compensate optically the chromatic dispersion in fiber using temporal Talbot effect in ranges exceeding the 40G bit/s. We propose a new optical post equalization solutions based on the self imaging of Talbot effect.

We have experimentally found that, for a communication system with many PDL and PMD components, the global PDL and the product of the maximum and minimum transmission coefficients (T_maxT_min) are wavelength dependent, even though they were treated as wavelength independent in the most of the simulations. At a given wavelength, however the PDL and PMD elements are concatenated, the global T_maxT_min of the system is equal to the products of the T_maxT_min of each comprising element.

Using an arrayed waveguide gratings (AWG) based demultiplexer, a simple channel gain equalizer can be designed. The gain equalization and blocking functions are realized by the hybrid waveguide based variable optical attenuators fabricated on the output waveguides of the demultiplexer. This paper discusses the operation principle of the design and presents some simulation results.

Use the method developed recently we calculate the bit-error-rate (BER) improvement as a function of transmitter extinction ratio and the optical link noise parameter.

In this paper, we investigate the effect of the fiber dispersion compensation and dispersion slope compensation in 160Gb/s single channel optical fiber transmission systems. Different dispersion maps are analyzed and their transmission performances are compared by calculating the Q factor. The numerical results show that there is no significant difference in the performance between the systems with and without dispersion slope compensation for the dispersion managed transmission systems when the launch power and the pre-compensation fiber length are all optimized. So, we conclude that the compensation for dispersion slope is not a must for dispersion managed fiber transmission systems. However, for the systems consisting of a single transmission fiber, the performance can be significantly improved by dispersion slope compensation. Moreover, the effect of local dispersion on the transmission performance is studied and the system performance improves slightly with higher local dispersion.

In this paper, we analyse the bit error rate reduction for return-to-zero coding with different duty cycle. It is shown that the optimal design of duty cycle of a given system lies in a compromise between spontaneous emission, intersymbol interferences and interchannel interference. We have carried out the semianalyitic simulations to find the dependence of optimum duty cycle on different system parameters. We found that the optimum duty cycle decreases as the channel spacing increases. For channels without ICI, BER is nearly a constant over a wide range of duty cycle as long as ISI is negligible.

Routinely used erbium crystal lasers operate at the 3-μm spectral range. In the silica fibers the transparency window corresponds to the eye-safe range of 1.5-μm. The ⁴I_13/2 -⁴I_15/2 transition provides the lasing at this range in a glass matrix. However, in crystals it is of negligibly small intensity. To significantly intensify this transition, the (Gd,Y)₃(Ga,Sc)₅O₁₂:Er³⁺ crystal has been chosen as the basis for the new laser crystal that is able to operate at 1.5-μm. The single crystal garnet films with thickness up to 18-μm were grown, using the method of liquid-phase epitaxy on the Gd₃Ga₅O₁₂ substrates. The 20-mol% maximal concentration of Er³⁺ ions was achieved without luminescence quenching. The up-conversion processes were blocked by the addition of the Fe-ions sensitizer. As a result, at the same level of absorbed pumping power the luminescence intensity at the 1.5-μm band for the specially doped film was approximately two orders of magnitude higher than that compared with the crystal of the traditional content. The spectral width achieved with a new medium is a little smaller than 300 nm, which makes this crystal convenient for the femtosecond laser design. The laser tunable inside this range may provide hundreds of the optical channels for telecommunication or optical computer devices.

C+L-band (1530 nm - 1605 nm) discrete fiber Raman amplifiers (dFRAs) with incoherent pumping are investigated considering gain, noise figure, and pumping efficiency with comparison to coherent pumping. It is shown that dFRAs with two incoherent pumping sources can have a gain flatness of less than 0.7 dB over 75 nm bandwidth. Noise performance of dFRAs with incoherent pumping is similar to coherent pumping. The comparison of C+L-band dFRAs with incoherent co-pumping to incoherent counter-pumping shows that dFRAs with incoherent co-pumping perform better due to not only lower but also flatter noise figure, besides increased pumping efficiency.

As the next generation of ultra-fast optical transmission systems, the optical time division multiplexing (OTDM) systems with a bit rate up to 160 Gb/s are actively being researched. Several experimental ultra-fast OTDM transmission systems with all-optical 3R-regeneration have been reported. With the development of ultra-fast OTDM technologies, techniques for rigorously evaluating the system performance through the calculation of bit error ratio (BER) have become increasingly important. We report a novel analytical model for estimating the performance of a 160 Gb/s OTDM receiver. The BER of the OTDM system is evaluated based on the calculation of the noise probability density function using the moment generating function. The optical pulse broadening by the fiber dispersion is compensated through the dispersion-managed approach and both nonlinearity and dispersion of the fiber channel are taken into account. Noise (ASE, shot and thermal) and performance-impairing factors (intrachannel interactions and timing jitter) are included in the calculation of the BER. A variational analysis approach is used to solve the optical pulse evolution over a periodical, dispersion-managed, nonlinear fiber channel as this yields an analytical expression for the received optical pulses. The OTDM demultiplexer is modeled as an optical gate controlled by an optical or electrical clock signal and the cyclostationary characteristic of the ASE noise after passing the OTDM demultiplexer is considered. Also the timing jittering between the signal pulse and the gate window and its effects on the signal decision are taken into account. Based on the proposed model, calculated results for the performance of the 160 Gb/s OTDM transmission systems are presented.

We present an analytical method to evaluate the effect of EDFA amplifier noise on the optical eye diagram for the system having PMD, PDL and CD in dynamic fiber links for any given optical pulse sequence. The method considers a single EDFA amplifier at the end of a fiber and considers all orders of PMD and PDL of the fiber. It gives the time dependent average output light intensity as well as its corresponding variations as a function of the amplifier noise, PMD, PDL and CD. Also the performance of the system is discussed for various parameters as well as the effect of OSNR on the system's Q-factor.

Extending the amplification bandwidth of erbium-doped fiber amplifiers (EDFAs) is one of the most cost-effective means of expanding the fiber transmission capacity. In conventional aluminosilicate EDFAs, gain drops sharply beyond 1605nm. Several new or modified erbium host materials have been used to extend the amplification band to beyond the conventional L-band, such as tellurite, bismuth-oxide, antimony silicate, P-doped aluminosilicate, and phosphosilicate EDFAs. Although tellurite and bismuth-oxide based EDFAs have wider bandwidths compared to that of demonstrated phosphosilicate (P-Si) EDFAs, P-Si EDFAs have been shown to provide better noise performance when the gain bandwidth is extended to 1620 nm. In addition, unlike tellurite or bismuth-oxide fibers, P-Si EDF does not exhibit increased nonlinearity or weakened reliability, compared to conventional aluminosilicate EDFs. Furthermore, phosphosilicate erbium fiber is compatible with other silica fibers and can be fusion spliced to the standard SMF silica fiber with high return loss and extremely low splice loss. These properties make the P-Si EDF a top contender for commercial extended L-band EDFAs. One key issue in the design of the extended L-band amplifiers is the optimization of power conversion efficiency while keeping the noise figure low. In this paper, we explore various amplifier configurations and compare their performances experimentally. We report high-power P-Si EDFAs with simultaneous improvement of PCE and NF by a combination of a single-pass low-noise stage with one or two double-pass high-efficiency stages. The best configuration yields high power (22dBm), low NF (5.5dB maximum over 1570-1620 nm band), and high efficiency (27% overall PCE after gain flattening).

Design of a high bit rate burst mode clock and data recovery (BMCDR) circuit for gigabit passive optical networks (GPON) is described. A top-down design flow is established and some of the key issues related to the behavioural level modeling are addressed in consideration for the complexity of the BMCDR integrated circuit (IC). Precise implementation of Simulink behavioural model accounting for the saturation of frequency control voltage is therefore developed for the BMCDR, and the parameters of the circuit blocks can be readily adjusted and optimized based on the behavioural model. The newly designed BMCDR utilizes the 0.18um standard CMOS technology and is shown to be capable of operating at bit rate of 2.5Gbps, as well as the recovery time of one bit period in our simulation. The developed behaviour model is verified by comparing with the detailed circuit simulation.

The demand for increased data and communications services is driving many carriers to upgrade their legacy optical systems in the metro-metro-edge region of the communication network. When such upgrades require extending the transmission distance, dispersion compensation techniques are required that compensate for signal degradations while conforming to tight operating budgets. We have used both channelized and non-channelized, fiber Bragg grating (FBG)-based dispersion-compensating modules (DCM) to compensate for the chromatic dispersion of 60-km fiber links operating at 1550 nm. Using typical SONET/SDH lasers with a large spectral linewidth (~1 nm at the 20-dB point), including significant chirp, our BER experiments show that the system performance depends on whether the grating-based compensators are channelized or non-channelized. Our measurements show that the performance of FBG-based channelized DCMs is a function of the amplitude of the dispersion ripple at the boundaries of the channel bandwidth. Non-channelized DCMs performed well despite their exhibiting more chromatic-dispersion and insertion-loss ripple in the immediate proximity of the laser center wavelength. First-order system simulation was found to be in general agreement with these experimental observations and singles out the large variations in chromatic dispersion at the boundaries of the channel as the primary reason for the degraded system performance.

The single-hop star-based network is proposed as a feasible topology to fit the rapidly-increasing bandwidth requirement in the AAPN research project. This paper investigates the node architecture to implement all-optical operations in such a network using available technologies. Based on the node placement in the network, two architectures are designed, one is placed in the edge and another one is used in the core. The edge node is a multi-stage electronic/optic switch, which aggregates legacy traffics and transmits them to the core node, or accepts optical messages from the core node and sends them to legacy networks. Each stage uses either electronic or optical components to implement signal storage, conversion or transmission. The core node is an all-optical switch which switches optical signals in different wavelength planes, while the controlling part works in electronic domain. A separate control plane is designed to manipulate the operation of different component devices. This system provides a common platform for the overlaid-star network. By introducing synchronization or not, we can employ reservation-based optical time-division-multiplexing (OTDM) or contention-based optical burst switching (OBS) in the designed architecture. No wavelength conversion or optical buffering is necessary by agilely scheduling the messages in both mechanisms. Our research is an efficient and feasible solution which satisfies the transmission requirement by taking into account of technological availability. Our design is supported by the performance evaluation of OTDM and OBS methods, and their comparisons under different scenarios.

The Routing and Wavelength Assignment (RWA) problem has attracted lots of attention in the research field for the past decade. Most of the existing works are the classic static RWA problem, which assumes every time for the reconfiguration, all the existing connections will be reconfigured. In a real operating network, the reconfiguration has to take the existing connections into consideration and any reconfiguration of the existing connection results in the disruption of the upper level traffic. The algorithms that are slow or do not consider the existing connections in the network cannot be used in the real-time reconfigurable network. In this paper, we propose the semi-dynamic/static network optimization problem that takes into consideration existing connections from the previous reconfiguration session. The objective function in the formulation is penalty-based, i.e., there are penalties for the reconfiguration of a connection, for the rejection of a connection demand and for the most congested link. Rules on the existing capacity and new demand in the new session are proposed. We have successfully used the Lagrange Relaxation (LR) and Subgradient Method to successfully solve this network optimization problem. This state-of-art frame work allows us to evaluate systematically some sample networks in terms of various network performances and behaviors. At the same time, excellent algorithm performance and efficient computation complexity are demonstrated.