Optical fiber sensor technology has progressed at a rapid pace over the last decade. Many different sensing techniques have been developed to monitor specific parameters. In particular, distributed Brillouin scattering-based sensor systems provide an excellent opportunity for structural health monitoring of civil structures by allowing measurements to be taken along the entire length of the fiber, rather than at discrete points, by using fiber itself as the sensing medium. One class of Brillouin-based sensors is based on the Brillouin loss technique, whereby two counter-propagating laser beams, a pulse and a CW, exchange energy through an induced acoustic field. This type of sensing has tremendous potential for structural health monitoring since the spatial resolution can be adjusted for different applications simply by altering the pulse duration, even after the fiber is installed. Although the spatial resolution can be improved using short pulse, the loss spectrum broadens as the pulse width decreases below the phonon lifetime t. Hence, it was generally believe that sub-meter resolutions were unachievable due to rapid linewidth increases when pulse width W < t = 10 ns provided a 1 m spatial resolution limitation. In this paper, we will report the development of distributed optical fiber sensor with centimeter spatial resolution. The sensing principle will be presented. We will also report the test results of pipeline buckling and corrosion fatigue monitoring and small damages/cracks of 1.5 cm in an optical ground wire (OPGW) cable with centimeter spatial resolution.

By combining a DC and a short pulse (~ 1 ns) laser as the probe beam in the pump-probe configuration of Brillouin-based fiber optic sensors, we have developed a new two-stage method to achieve both high accuracy strain/temperature measurement and spatial resolution (order of cm ). Brillouin profile of such a configuration has a two-fold structure consisting of a Lorentzian- and a Gaussian- like portions. Lorentzian-like portion, although gathering information from all over the fiber, has localized and position-dependent information that can be extracted and employed in a signal processing method for high accuracy and resolution measurements. Level of DC to pulse power affects both the position-dependent information of the Lorentzian peak and its SNR relative to Gaussian peak. An optimum level of DC to pulse power for best SNR and position detection capability is discussed.

High sensitivity, real time distributed and cost effective sensor system is in great need for structure healthy monitoring in civil engineering. In our lab, we are developing a distributed, Stimulated Brillouin Scattering based, fiber optic sensing system at 1550nm wavelength. Our current SBS-based fiber optic sensor system works at 1310nm wavelength. Two expensive Nd: YAG Lasers (US$40,000 each) are being used, which leads to a soaring high cost to the entire system and eventually limits its application. Distributed Feedback (DFB) lasers have large tenability, compact size and low cost (less than US$1000 each). But they are not stable enough for the sensing system. In this project, we use the frequency offset locking technique with optical delay line and electrical feedback circuit to optimizing the stability of DFB lasers so that the lasers of 1310 nm in the sensor system can be substituted by the lasers of 1550 nm that is the most often used band in modern fiber optic telecommunication system. Less than 100 kHz stability of the beat frequency is required to achieve temperature accuracy of 0.1°C and strain accuracy of 2me. In our system we have realized 20 kHz stability of beat frequency of two DFB lasers. Greater than 800MHz turning range is necessary for the detection of temperature range of 600 °C and strain range of 10,000 me. In our system we have achieved 925 MHz in 18.75 seconds. In the sensing part, we can vary the pulse width from 120ns to 5ns that means we can realize the spatial resolution of 50cm at least. Because the total optical loss in the setup is comparably smaller, the measurable fiber length is mainly determined by the optical power launched to the fiber, normally it is in tens kilometers.

With the rapid development of distributed fiber optic sensor, such as Brillouin Optical Time Domain Reflectometry (BOTDR), much more importance is attached to the performance of optical fiber. As a conductive and sensing element, optical fiber must be protected from damage and meanwhile be sensitive to strain and temperature. Therefore, tight-buffered optical fiber with a jacket is often used in a certain harsh environment instead of ordinary optical fiber. The jacket can protect optical fiber, on the other hand, exert effect on the optical fiber's sensitivity to temperature and strain. In this paper, two types of optical fiber, i.e. with jacket (Type-A) and without jacket (Type-B), were selected to study the jacket effect on strain measurement, and relative experiment schemes and results are present. It is found that the strain of Type-A, measured by BOTDR, changed with time under constant load stretching within 48 hours, while the Type-B kept constant. After that the load was removed to let the fiber relax. After several days, then again loaded and measured. Through three cycles, the strain measurements of the Type-A gradually increased, but the accretions become smaller and smaller. These experiment results indicate that the jacket has the creep deformation that effects the strain measurement.

Results of research on real-time distributed monitoring of agents and products of materials deterioration will be presented. The potential of integration of functional polymers and optical wave-guides as tools for monitoring materials chemistries will be discussed. In such hybrid sensing systems, analyte recognition is carried out through the selection process inherent in the fiber optic poro-selective cladding materials, while transduction is carried out by the guided evanescent field which is spectrally modulated according to the level of the target analyte species present.

Pipeline failures induce costly repair and cleaning spending that could be avoided by the implementation of proactive approaches. Distributed sensors based on Brillouin scattering are attractive candidates to monitor structural health of pipelines. They can measure local strain and allow real-time control over lengths ranging from a few meters to tenths of kilometres. One of the possible degradations that must be detected is buckling. We describe in this paper what is to our knowledge the first time report of buckling detection with a Brillouin sensor. We conducted an experiment reproducing buckling monitoring in a laboratory environment. Two specimens (steel pipeline and beam) were prepared by locally thinning the inner wall to provoke buckling. Fibre was laid along the external walls of the specimens. Strain gauges were glued in thinned wall area. An axial load was applied to the specimens and increased while strain measurements were carried out with the Brillouin sensor and the strain gauges. All the measurements showed a progressive compression increase in the neighbourhood of the thinned wall. Finally buckling aroused and was visually identified as well as localized with the Brillouin sensor. Strain gauges readings and strain measurements with the Brillouin sensor were in good agreement.

Presented are long-period gratings (LPGs) fabricated in pure silica photonic crystal fibre (PCF) using an electric arc. Two different varieties of PCF have been investigated, an endlessly single mode PCF and a large-mode area PCF. The LPGs have been characterised for their sensitivity to a variety of external measurands. The LPGs in both fibres have been found to have negligible temperature sensitivity whilst exhibiting good sensitivity to bending and strain.

This paper reports the results of our recent investigation on the noise limited performance in heterodyne interferometric demodulation systems for fiber Bragg grating strain sensors. Theoretical and simulation results are presented and compared with experimental results.

The air holes structured in a Photonic crystal fiber’s cladding are easily contaminated by dust and moisture. For some applications, the fibers need to be in contact with fluids. The capillary effect will then draw the fluid into the air holes and change the optical properties of the fiber. A simple solution to avoid this effect is to collapse these air holes near the fiber end face. This, however, will in turn significantly affect the optical properties in such a way that they will need to be specifically investigated. In this paper we present a theoretical model of light transmission in the area of a collapsed fiber end face. We demonstrate that the air hole collapsed PCF could be represented by an equivalent PCF fiber with complete holes. A shorter collapsed segment will lead to a higher accuracy of this model and to a higher power coupling efficiency than a longer collapsed segment. Error distributions are calculated for different incident angles and for different core sizes. A 0.7 m length of a PM-1550-01 polarization maintaining PCF fiber is experimentally investigated. The setup uses an optical fiber splicer to couple light from another PM-1550-01 PCF fiber to the investigated PCF fiber. A SLD is used as a light source. The transmitted light is measured by a EXFO Fibre-Optic Tester. By adjusting the distance between the launching fiber and the receiving PCF fiber (10 mm per step), the intensity-distance curves are generated for 3 different collapsed distances: 1. Air holes of PCF fiber are not collapsed; 2. Air holes are collapsed at about ~1/5 OD and 3. Air holes are collapsed at about ~1 OD. The analysis of these curves confirms the validity of the developed theoretical model.

A high survivability fibre sensor network is presented which has applications in smart structures. Linear sensor arrays suffer from the weakness that a fracture in the array will cause the loss of information from all sensors beyond the fracture. This scheme illustrates a method that is capable of withstanding damage with little or no effect to the data collection efficiency. Four different wavelength FBGs were fabricated in single-mode optical fibre and spliced to two 50:50 couplers. The reflected signal from the sensor network was observed on an optical spectrum analyser. Points along the network were attenuated by 100% to simulate a break. The attenuation of two points on the arms showed little effect to the reflection spectrum. Attenuation to four other points showed the loss of the reflection from only one grating whilst maintaining the signals from the remaining gratings. This compares favourably to a standard fibre sensor array that would lose one or more gratings, depending on the point of attenuation. Such a system is highly desirable in rough operating environments such as found in the military or in heavy industry. The system introduces a level of redundancy without the expensive need of duplication. Other techniques involving switches or higher multiplexed couplers may be used but they present their own disadvantages. Optical switches are expensive and not particularly rugged and the higher multiplexed coupler technique has a power penalty as the power is split over a larger number of arms.

We analyze the polarization properties of uniform fiber Bragg gratings written into polarization maintaining fibers by studying the evolution with wavelength of the normalized Stokes parameters and the degree of polarization for both the reflected and transmitted signals. Theoretical expressions are derived for the Stokes parameters. Numerical simulations and experimental results are also reported. We demonstrate the possible realization of a sensor using the information contained in the evolution of teh degree of polarization in reflection. This new demodulation technique has been experimentally tested with temperature sensors.

Bragg gratings have widespread applications in the rapidly growing field of optical sensors. Although fiber Bragg gratings are often referring to permanent refractive index structures, exposure in increased temperature usually results the decay of the refractive index modulation. Basically, the stability of the grating competence at high temperature is an important criterion for high temperature sensor applications. This report is a part on going research to develop high temperature optical sensors. We report our design and analyze of a hydrogen loaded fiber Bragg grating temperature sensor range from room temperature to around 1000°C. A basic setup has been built in our lab to examine the performance of the point temperature sensor based on the hydrogen loaded fiber Bragg grating. Until now, a grating has been shown to stabilize at temperatures in excess of 700°C and to survive at temperatures in excess of 930°C. The tested operation gratings around 700°C retain up to 80% reflectivity after one and a half hours. The thermal treatment of the tested hydrogen loaded fiber Bragg gratings is demonstrated capable to enhance effectively the grating's thermal stability. Our experimental results provide a better understanding of thermal response to the hydrogen loaded fiber Bragg gratings and their decay behavior at elevated temperatures.

The inscription of fiber Bragg gratings during the drawing process is shown to be a very useful method to realize sensor arrays with high numbers of gratings and excellent mechanical strength and also type two gratings with high temperature stability. Results of single pulse grating arrays with numbers between ten and >100 and definite wavelengths and positions for sensor applications were achieved at 1550nm and 830nm using new photosensitive fibers developed in IPHT. Single pulse type-1 gratings at 1550 nm with more than 30% Reflectivity were shown first time to our knowledge. The mechanical strength of this fiber with an Ormocer coating with those single pulse gratings is the same like standard telecomm fibers. Weibull plots of fiber tests will be shown. At 830 nm we reach more than 10% reflectivity with single pulse writing during the fiber drawing in photosensitive fibers with less than 16 dB transmission loss. These gratings are useful for stress and vibration sensing applications. Type two gratings with reflectivity near 100% and smooth spectral shape and spectral width of about 1 nm reach temperature stability above 1200 K. They are also realized in the fiber drawing process. These Gratings are useful for temperature sensor applications.

A displacement sensor has been newly developed using a hetero-core spliced fiber optic element in a form of simple module structure and evaluated in terms of the accuracy, reproducibility, and sensitivity as an OTDR-based, tandem connected displacement detector, for the purpose of the full-scale environmental monitoring. The developed sensor module is designed to be sensitive to a relatively large displacement in the range 0-5mm, for which a simple displacement-macrobending conversion mechanism in the module produces the change in the sensor transmission loss ranging from 0 to a few dB. The initial insertion loss only due to a hetero-core portion was 0.3dB. The sensitivity showed a tendency proportional to the insertion length of hetero-core portion in the range from 1 to 2mm. In contrast to conventional fiber distortion sensors such as FBG (Fiber Bragg Grating) and BOTDR (Brillouin Optical Time Domain Reflectometer), the hetero-core technique introduced in this work shows no temperature dependence in its sensor principle since the use of hetero-core makes it possible that change in backward Rayleigh scattering can be successfully amplified with a relatively large curvature. The experiment promisingly showed sufficient reproducibility the sensing operation with the accuracy less than 0.1% to the full span displacement of 5mm.

We analyze the crosstalk of compliant cylinder type fiber optical accelerometers using 3D finite element analysis. For a standard design we obtain vibrational modes (transversal and torsional) having their eigenfrequencies below the eigenfrequency of the principal mode, i.e. the longitudinal mode vibrating parallel to the axis of axial symmetry. Through the inclusion of two diaphragms we show that for this new topology the principal mode is the fundamental mode. We have constructed such topology and tested it. A responsivity of 2000 rad/g at an optical wavelength of 1550 nm and a crosstalk of -26 dB was achieved, ignoring one resonance.

Laser Diodes have been employed many times as light sources on different kinds of optical sensors. Their main function in these applications was the emission of an optical radiation impinging onto a certain object and, according to the characteristics of the reflected light, some information about this object was obtained. Lased diodes were acting, in certain way, just as passive devices where their only function was to provide the adequate radiation to be later measured and analyzed. The objective of this paper is to report a new concept on the use of laser diodes taking into account their optical bistable properties. As it has been shown in several places, different laser diodes as, for example, DFB lasers, offer bistable characteristics being these characteristics a function of different parameters as wavelength, light polarization or temperature. Laser Bistability is strongly dependent on them and any small variation of above parameters gives rise to a strong change in the characteristics of its non-linear properties. These variations will be analyzed and their application in sensing reported. It will be reported the dependence on wavelength, spectral width, input power and phase variations, mainly for a Fabry-Perot Laser structure as basic configuration.

The combination of a Malus and Fabry-Perot interferometers using fiber optic devices has been proven useful to achieve an enhancement in sensitivity to measure changes in circular birefringence. This fiber optic Malus Fabry-Perot interferometer (FOMFPI) allows for the sensitive detection of changes in the polarization of the guided beam due to torsion applied to the fiber, owing to multiple passes of the beam through the sensing area. We present a theoretical analysis based on Jones calculus showing that it is possible to measure variations in circular birefringence of the fiber upon registering changes in the transmission through this arrangement. The matrix representation developed for the FOMFPI allows for the evaluation of its performance as a function of parameters such as mirror reflectivity and intra-cavity losses. Experimental results using both, bulk optical components and optical fibers, show that compared to single-pass polarimeter measurements, enhancement in sensitivity is effectively achieved. The dependence of the enhancement factor on the reflectivity of the mirrors is evaluated upon using mirrors with variable reflection coefficient. The performance of FOMFPIs using multimode and single-mode fibers is also investigated.

Polarization sensitive optical coherence tomography (PS-OCT) takes into account the vector nature of light waves (state of polarization). The most complete information about the polarization properties of a biological target is given by the depth resolved Mueller matrix elements, however, it is difficult to construct such a system for in-vivo examination. We designed and assembled a simpler system, with two incoherent channels to provide limited information, but essential on the polarization properties of the tissue. The interferometer is a hybrid configuration of bulk optic and single mode optical fiber components. No polarization maintaining fiber is used. The reference and sample beams interfere in single mode optical couplers. The low coherence light source is a superluminescent diode of 850 nm center wavelength and 25 nm spectrum FWHM (which corresponds to a depth resolution of 12 microns in tissue). The system can display either a pair of two polarisation sensitive OCT images, corresponding to linear orthogonal polarisation directions or a pair of images, a polarisation insensitive (pure reflectivity) image and a birefringence retardation map. The 12 bit grayscale images are collected by fast en-face scanning (T-scan) at 2 frames/s. We demonstrate in vivo en face images of the retinal nerve fiber layer, lamina cribrosa, cornea and teeth. A rotation angle of 0.3 degrees per micron was evaluated from the retinal nerve fiber layer and lamina cribrosa.

Design and characterization of a pH optical fiber sensor with a pH sensitive dye is described in this paper. TEOS (Tetra-ethyl-Orto-Silicate) was used to dope a plastic optical fiber, which will be used as the probe. The sensor is prepared by fixing the doped plastic fiber on a fused ortosilica block surface with blue bromophenol. The fiber surface charged with silica and the refractive index, which plays an important roll on the fiber, modifies the conditions of light propagation into the plastic optical fiber. The fiber transmittance is used to measure the pH of a solution or a fluid in a range between 7 and 10.5. Such signal is captured by a photodetector and processed with a LabView development environment which also controls a hardware designed in our laboratory. The advantage of this system is that 2cm of doped fiber are enough to measure the pH of a fluid in real time. The response time of this system is approximately 10s.

Sensors for the detection of hydrogen, which is very volatile, extremely flammable and highly explosive, are needed in many fields, for example in nuclear power plants, in launch vehicles for human space flight, in hydrogen production plants, and fuel cells. To enhance the safety level it seems appropriate to use optical sensors instead of electronic ones. In the presentation, investigations concerning the use of fiber optic Bragg grating (FBG) sensors as part of the surveillance system of liquid fuel tanks will be described. The sensors are supposed to measure strain and temperature when embedded in the inner tank wall and to detect hydrogen leakage. To fulfill this task the temperature and strain sensors have to prove their functionality down to minimum temperatures of 20K. As the dn/dT for quartz decreases to very low values, the FBG temperature transducer has to be bonded to a substrate with sufficiently high coefficient of thermal expansion also under cryogenic conditions. As the only applicable solution a high-expansion glass has been found, which in connection with specific fiber coating and bonding materials for the strain-transducer fulfils all the sensor requirements. The hydrogen sensor utilizes the expansion of palladium in an atmosphere containing hydrogen and consists in our new sensor configuration of a palladium foil bonded to a special-shaped fiber. Experimental results for concentration ranges 0-4%vol H2 and temperature ranges from -40°C to +80°C show the parameters of application of a hydrogen leakage detection system.

The control of environmental conditions, such as temperature, pressure, and humidity, are important in many applications ranging from bio-medical to space exploration. Proper humidity control is also important in the conservation of organic materials. Therefore an accurate and sensitive method to characterize the moisture content of the particular environment is of valuable importance. This paper proposes a humidity sensitive polyimide material as a fiber optic sensor for humidity measurements. The spectral analysis and the intensity of transmitted light through the polyimide sensor will represent the humidity measure. The paper also presents the absorption characteristics of the proposed humidity sensitive material. The experimental values on the spectral shift and light intensities are measured at different humidity conditions. This paper will also present the feasibility study for using the proposed fiber optic sensor for humidity measurements.

The study and measurement of two-phase flow is very important in industry, since the components, contents and distributions of the flow have great influence on the produce involving its quality, cost and even safety. Many methods have been perused including electrical capacitance tomography (ECT), electrical impedance tomography (EIT) and ultrasonic computed tomography (UCT), while optical fiber process tomography (OFPT) is a new one, which owns all the advantages of the optical fiber sensors, such as small, safety, free of electromagnetic interference and having high sensitivity in the measurement of low-density transparent media. In this paper, a practical optical fiber process tomography system is developed, which can be used in the industry filed to measure the medium distributions and contents. The liquid-solid samples are measured and reconstructed and the results show that positions of the solid objects are reoccurred exactly. The crude oil with opaque performance can be regarded as the solid, so that the oil-water two-phase flow can be measured the same as liquid-solid two-phase flows by the new OFPT system. The relative reconstructed resolution is 1.67 from experimental results.

It is very important for hydrofluoric acid to improve the performance of fused optical fiber and to prepare the fiber sensor. A novel transmitted-light differential interference contrast (DIC) system was used to nondestructively measure the refractive index profile (RIP) of an optical fiber. A hydrofluoric acid solution effectively enhanced some performances after fusion occurring. Furthermore, it also can remove the cladding of a optical fiber for constructing a fiber sensor to monitor corrosion of steel and aluminum.

Brillouin scattering-based distributed fiber optic sensors have been shown to be effective diagnostic tools for monitoring structural health, and detecting fires and hot spots, among other uses. Current research has mainly been focused on improving the spatial, strain and temperature resolutions, and sensing lengths of these systems, generally by the use of better signal processing and improved equipment. In contrast, there has been little published work on optimizing the sensing optical fiber itself. A number of commercially available optical fibers have been measured in order to determine how to optimize their Brillouin characteristics. Some characteristics chosen are the number of Brillouin peaks, the frequency of the peaks, their linewidth, and the temperature and strain coefficients of each peak. It is shown that lowering the intrinsic Brillouin frequency of the fiber can increase the Brillouin strain coefficient and decrease the temperature coefficient of the optical fiber for the main Brillouin peak, among other results.

Timely, accurate and reliable pressure information about how the reservoir is performing is an important component to optimizing oil yield and production rates. This paper reviews the use of fiber optical pressure sensor for downhole monitoring in the oil industry. Several types of pressure transducer with different characteristics have been introduced. Due to their multiplexing capabilities and versatility ,the use of Bragg grating sensors appears to be particularly suited for this application. A sensor for accurate and long term fluid pressure monitoring based on optical fiber Bragg gratings(FBGs) is developed. The sensor converts fluid pressure into optical fiber strain by means of a mechanical transducer to enhance its sensitivity to pressure. It can also implement distributed or multiplexed sensing. The sensor operation is studied at pressure up to 100 Mpa(1000bar) and the tested temperature to ~175°. It is possible to be used in the well.

Distributed Sensors based on Brillouin scattering are attractive candidates to monitor structural health of mechanical structures. Currently physical limitations of time resolved techniques do not allow extracting accurate strain when pulsewidth is larger than strained length. The purpose of this study was to quantify the errors in such a case and derive a decision threshold making possible the discrimination of strained section from unstrained contribution. We achieved these goals by solving the coupled intensity equations for various strained lengths, pulsewidths, strain strengths, Brillouin linewidths, Fibre lengths, probe and pump powers. We observed two dominating behaviours principally functions of strained length and strain strength. In the first regime, a single peak broader than Brillouin natural gain/loss peak was detected. In the second case, we could distinguish two peaks. From these behaviours we then derived curves relating Brillouin peak contrast as well strain accuracy to the ratio between strained section length and pulsewidth. These curves also show a strong dependence on strain amplitude. This led us to find a strain induced frequency shift threshold characterising the transition from the single peak regime to the double peak regime. Finally we defined a Rayleigh equivalent criterion that unambiguously separates strained from unstrained contributions.

The optical fiber EFPI sensor with cavity length shorter than 20 µm is suited for structure health monitoring. The cavity length of EFPI is crucial measurand and usually obtained by low coherence interference method, but the coherence lengths of most laser source are greater than 20 µm, which lead to the overlapping of multiple interference fringes and failing to obtain cavity length. In this paper, a new demodulation method is proposed. The EFPI sensor of system is made with multimode fiber, the other part is monomode fiber, and a low coherence interference experiment setup with ASE light source, which avoids overlapping of different orders of low coherence interference fringes wave packet through reflectivity match, is established. A algorithm is developed for retrieving cavity length and a new method for multiplexing of EFPI sensors is presented. Experiments show that the cavity length of EFPI can be obtained efficiently and multiplexing is easily realized.

We solved the three coupled partial differential equations in transient regime for the probe-pump Brillouin sensor to explain the sub-peaks in Brillouin loss spectra, which have been experimentally observed. We discovered that the Fourier spectrum of the pulsed signal and the off-resonance oscillation attributed to sub-peaks. The off-resonance oscillation at frequency [v - v_B] is the oscillation in the Brillouin time domain when the beat frequency v of the two counter-propagating laser beams does not match the local Brillouin frequency v_B. This is important in differentiating the sub-peaks from strain/temperature peaks.

An optical preamplifier system consisting of a 27dB-gain erbium-doped fibre amplifier, a three-port circulator and a 47GHz bandwidth in-fibre grating has improved receiver sensitivity and consequently sensor range of a spontaneous Brillouin intensity based single-ended fiber-optic distributed temperature sensor. Optical signal-to-noise has improved by 17dB. Operating at 1533nm, a sensor with a range of 23km, spatial resolution of 1.8m, measurement time of 9 minutes and temperature resolution of 6K was implemented. The temperature resolution was attributed to contamination of the backscattered spontaneous Brillouin signal by the residual Rayleigh signal after optical filtering. The optical signal-to-noise is supported theoretically by considering asscociated noise sources.

Optical fiber laser sources with switching capabilities are of interest in applications such as WDM fiber communications systems and optical fiber sensors. In particular, a polarization-switched fiber laser would be a suitable light source for optical fiber sensors linked through an optical bridge that could lead to the development of compact and low noise sensing configurations. We have studied the performance of a fiber laser in which polarization switching is achieved through intra-cavity devices. As a first approach, we have used bulk optical components within the laser cavity, and switching between two orthogonal polarization states is achieved by adjusting the coupling between the polarization modes and by independently varying the losses for each mode. An all-fiber version of the proposed laser configuration was realized using an intra-cavity electro-optic polarization switch. Upon comparing the performance of both configurations we have evaluated the effects of intra-cavity losses and length of the gain medium on switching speed of the fiber laser. Whereas only low switching frequencies (60 Hz) are attained with the first configuration, the all-fiber arrangement yields switching frequencies limited by the relaxation oscillations of the fiber laser. Thus, we demonstrate a simple arrangement for a polarization switched fiber laser using commercially available optical fiber devices.

Current software simulation tools in integrated optics allow to tackle directly a multitude of complex problems which occur in high index contrast waveguiding. Appropriate eigenmode solvers and propagation schemes account for the vectorial character of the fields, and transmissive and reflective device properties can be derived. For the fibre-chip coupling of high index contrast waveguides the field matching requires to decrease the mode diameter by more than an order of magnitude. If photonic crystal waveguides are considered, the match of impedance and phase additionally call for attention. Generalized coupling integrals based on the modes at both sides of an interface take all these effects into account. This allows for an efficient analysis of conventional waveguide to photonic crystal waveguide interfaces. Moreover, free-space to waveguide coupling can be analysed as well, if an appropriate field transfer from vectorial raytrace to wave propagation is implemented. Thus, the design of hybrid mode transforming optics for fibre-chip coupling including micro-lenses and integrated waveguide taper structures can be addressed.

This paper presents the transmission line matrix method (TLM) as an alternative efficient simulation tool for the analysis of photonic crystals (PCs). The paper describes important aspects for the computation of the photonic band structures of infinitely periodic PCs within the formulation of the TLM method. In addition, we propose two methods for reducing the computational effort involved in the simulations of PCs. One method is based on a real-valued implementation of the periodicity (Bloch) condition, and the other one is based on the use of a multi-grid mesh. Depending on the physical geometry of the crystal, computational savings of over 50% can be easily achieved. The advantages and limitations of these methods are described. Given the popularity of the finite differences time domain (FDTD) method for the simulation of PCs, we briefly compare the performance of the TLM method with that of the FDTD and show that under various circumstances, the use of the TLM method can be advantageous. The suitability of the TLM method to handle PCs with more general material properties such as frequency dependent metals and semiconductors is also demonstrated. Finally, we validate these simulation aspects of the TLM method by simulating various photonic crystals composed of dielectric, metallic and semiconducting materials using uniform and multi-grid meshes. The results are compared with those predicted by alternative methods such as the plane wave expansion method for verification.

Optical scattering loss in sub-micron scale patterned waveguides is one of the most important physical mechanisms dictating the limitations and applications of optical devices containing such structures. Despite this,there has been little theoretical work describing the extrinsic scattering losses in photonic crystal waveguides due to random fabrication variations such as disorder and surface roughness. While much work has been devoted to the characterization of ideal, lossless photonic crystal devices, the role of extrinsic optical scattering loss has not yet been suitably addressed. We present explicit formulas that describe extrinsic optical scattering loss for arbitrary sub-micron patterned waveguides occuring due fabricated imperfections such as disorder and surface roughness. Using a real-space Green function formalism, we derive original expressions for the backscattered loss and the total transmission loss, including out-of-plane contributions. Numerical calculations for planar photonic crystal waveguides yield extrinsic loss values in overall agreement with experimental measurements reported in the literature. Additionally, our formulas offer physical insight, including scaling rules that indicate how waveguide losses may be reduced by improved design. In particular, we highlight that loss becomes unavoidably large for operating frequencies approaching the photonic bandedge.

A transmission-line laser model has been used for simulating distributed-feedback (DFB) lasers. Statistical distributions of laser parameters like threshold current, slope efficiency, front-to-back power ratio, or side-mode-suppression ratio (SMSR) are generated by varying randomly lasers’ facet phases. Model parameters were adjusted by comparing simulated and experimental distributions for a continuous wave (CW) index-coupled and a 2.5 Gbit/s gain-coupled directly-modulated (DM) DFB lasers. For the index-coupled DFB laser, agreement with experimental data is excellent except for the front-to-back power ratio, which has a larger spread than measured experimentally. For the gain-coupled DFB laser, distributions are in excellent agreement with experimental data, but the SMSR is calculated to have a median about 6 dB larger than measurement. Distributions of small-signal parameters and dispersion penalties after propagating in an optical fibre are also generated for various drive conditions and design parameters. It is shown that a grating with an index coupling larger than 4.0 and a gain coupling of around 0.05 gives the highest 2 dB dispersion penalty yields for a reach of 450 km. There is nevertheless a compromise between high dispersion penalty yields and CW single-mode yields when using large index coupling coefficients with only a small amount of gain coupling.

In our recent published works we reported that the propagation length dependence of the conversion efficiency of the Cerenkov second harmonic generation (CSHG) in planar waveguides with a nonlinear substrate follows quite strict rules depending on an actual waveguide geometry together with the wavelength of the pump radiation. Namely, simple integral expressions show that the propagation length exponent may vary continuously from zero to a quadratic dependence. In this contribution, we generally analyze the propagation length dependence of CSHG for some specific arrangements. We also study, for the first time, the effect of pump depletion together with the influence of the pump wavelength and the effective refractive index. It is shown that the length dependence can thus be described by several categories, where the conversion efficiency is - with respect to the propagation length - linearly proportional and quadratic, but also the length exponent may continuously vary from 1 to 2 (for the classical phase-matching), including the so-called Cerenkov peak region, where the length exponent is equal to 3/2.

Diffractive optical elements (DOE) utilize diffraction to manipulate light in optical systems. These elements have a wide range of applications including optical interconnects, coherent beam addition, laser beam shaping and refractive optics aberration correction. Due to the wide range of applications optimal design of DOE has become an important research problem. In the design of the DOEs, existing techniques utilize the Fresnel diffraction theory to compute the phase at the desired location at the output plane. Since this process involves solving nonlinear integral equations, various numerical methods along with robust optimization algorithms have been proposed. However all the algorithms proposed so far assume that the size and the spacing of the elements as independent variables in the design of optimal diffractive gratings. Therefore search algorithms need to be called every time the required geometry of the elements changes, resulting in a computationally expensive design procedure for systems utilizing a large number of DOEs. In this work, we have developed a novel algorithm that uses neural networks with multiple hidden layers to overcome this limitation and arrives at a general solution for the design of the DOEs for a given application. Inputs to this network are the spacing between the elements and the input/output planes. The network outputs the phase gratings that are required to obtain the desired intensity at the specified location in the output plane. The network was trained using the back-propagation technique. The training set was generated by using genetic algorithm approach as described in literature. The mean square error obtained is comparable to conventional techniques but with much lower computational costs.

The National Research Council's Herzberg Institute of Astrophysics (NRC-HIA) has developed an opto-mechanical integrated modeling toolset called TM-IM. This time-domain state-space toolset has been implemented using Matlab/Simulink/C. The toolset was originally developed for the Very Large Optical Telescope (VLOT) design work, and continued when Canada joined in the Thirty Meter Telescope (TMT) project. The TM-IM toolset has been developed to accommodate different structural and optical designs and has been used to evaluate telescope performance to assist in making decisions for the TMT reference design expected fall 2004. Preliminary results include delivered image quality as a function of wind loading on the structure, primary and secondary mirror, and the simulation of an Adaptive Optics system which provides control feedback to the primary mirror.

As optical and illumination systems increase in complexity, it is important that the tools used to design and analyze these systems provide better and more efficient methods for the engineer to correctly model the systems to achieve the most accurate results possible. Important considerations include better interoperability between different analytical tools sharing the total calculation as well as providing more robust interchange between the CAD and optical environments. Several new features in the Advanced Systems Analysis Program (ASAP) are directed at achieving these goals.

In this paper we investigate the limits on optical detection of noisy watermarks that use a chaotic function, the logistic difference equation, in the watermark generation scheme. By varying the function seed, different chaotic sequences exhibiting lowpass and highpass characteristics, can be obtained for the same function, offering an added security advantage over watermarks generated using pseudorandom sequences. Watermark Detection is the process of determining whether an image is watermarked with a certain watermark. In this paper, we model and investigate an optical correlator suitable for watermark detection for certain classes of high-pass or low-pass watermarks. Once in the public domain a watermarked image may be subjected to noise and other attacks, deliberate and unintentional. Additionally,an optical correlator system will also be subject to shot noise. The effects of shot noise on optically transmitted watermarks are modeled in this paper and we examine how the watermark detection scheme performs in such situations. We quantify the degree of noise that may be present in the watermark detection scheme in order to obtain reliable detection or rejection of a watermark using an optical-correlator.

Simulation has become central to the successful development of integrated optoelectronic components and devices. Due to its flexibility and ease of use, the Beam Propagation Method, BPM, has established itself as one of the most popular and useful modeling techniques currently available. Many versions of BPM have been explored and presented in the literature with various schemes used to discretize the transverse operator. Vector and wide angled formulations as well as bi-directional schemes have been shown to overcome the inherent scalar, paraxial and one-way propagation assumptions associated with the simpler schemes, consequently expanding the range of practical problems for which BPM is suitable. However, there remain many significant practical scenarios to which BPM currently has limited applicability, for example it struggles to cope with structures in which there are many reflections or with physically large geometries where stringent performance specifications for new designs demand highly accurate 3D simulations. We assess these limitations in comparison with other simulation techniques and consider the major improvements required in the near future.

Finite difference beam propagation modelling (FD-BPM) calculates the effect of thermal gradients in an optical backplane on the optical field propagation in a polymer multimode linearly tapered waveguide. Compared to straight waveguides, tapered entrances offer improved power coupling for a wide range of optical source lateral offset misalignments. However, surface temperature gradients of 0.5°C/μm across the taper were found to degrade this benefit due to the thermo-optic effect of the polymer. Higher surface temperature gradients improve power coupling in two discrete ranges of lateral source offset but the original tolerance to a wide range of source offsets is not recovered.

A numerical transmission line model for semiconductor lasers with long external fibre cavities and fibre grating reflectors is reported. The model is applied to high-power 980 nm lasers and shows good agreement with experiments in the coherence collapse regime (CCR). The numerical approach has been used to model the saturable absorber in the external cavity of the semiconductor laser with the external fibre grating in the "coherence enhanced regime" (CER). Our simulations show for the first time how line-narrowing evolves in CER lasers.

Microstructured optical fibers (MOFs), including Holey and multi-layered fibers have attracted great interest both in applications and theory due to their wide range of novel optical properties particularly adjustable nonlinearity. Up to date, analytical/numerical models have been developed to determine the electromagnetic fields distribution in the transverse directions and for various MOFs. We have developed a general three dimensional (considering cylinderical symmetry) analytical-numerical model of Stimulated Brillouin scattering in MOFs in which both electromagnetic field distribution in the transverse direction and its propagation due to a nonlinear effect (Brillouin scattering) is studied. The model has been employed to describe the Stimulated Brillouin Scattering phenomena in single and multi-mode core-cladding fibers. We examine how the structure of such a fiber like core size, index profile, and the laser wavelength affect the Brillouin profile in single and multi-mode regimes. We refer to a specific application of Brillouin scattering in fibers i.e. Brillouin based fiber optic sensors and specify the parameter space (core size, index profile, wavelength) for optimum sensing capability.

EDFAs have been used for some years now in building effective long-haul optical systems for the most diverse applications. For some applications, it is necessary to introduce some feedback control laws in order to avoid the generation of transients that could generate impairments in the system. In this paper, we use a multi-objective optimization approach based on genetic algorithms, to study the introduction of PD controllers into systems of cascaded EDFAs. We compare the use of individual controllers for each amplifier to the use of controllers to sets of amplifiers, and develop guidelines that help designers in making decisions while they are developing their own case-specific designs.

Variable optical attenuators (VOAs) play an important role in the wavelength division multiplexed (WDM) telecommunication networks. It is often necessary to use VOAs to perform functions as: (a) Dynamic channel balancing at MUX location; (b) Dynamic channel leveling at add/drop sites; (c) Receiver overload control; and (d) Optical channel blocking. In order to minimize the overall costs and optimize the performances, VOAs are sometimes required to be integrated with other optical components, such as a MUX/DEMUX, a detector array and etc. Planar lightwave circuit (PLC) technology is the ideal technology platform for realizing this type of large-scale integrations. In this work, we propose a VOA design based on a simple straight polymer channel waveguide layout in order to provide the optimal optical performances with very low electric power consumptions and high fabrication yield. The fabrication of the proposed design will be in compatible with other polymer components on the same substrate. It is found from the simulation results that the design can offer high attenuation level with very low electrical power consumption. In addition, the simulation results demonstrate that the proposed VOA design can achieve very low polarization dependent loss and very good spectra flatness. All these performances would make the proposed VOA design very suitable for large-scale integration applications.

Stimulated Brillouin scattering (SBS), in an optical fiber, is a three-wave interaction (3WI) resulting from a coupling between light and acoustic waves. In a fiber optic sensing context, SBS results from the interaction between counterpropagating pulsed and continuous fields. We formulate a solution to the time dependant, one dimensional 3WI model in a SBS based fiber optic sensor. It is shown that a low complexity, first order finite difference time domain (FDTD) solution is capable of accurately modelling the dynamics of SBS with little computational effort. A modification to the first order scheme is proposed to combat numerical damping and dispersion, brought on by the low order of the solution. Examples are presented, validating the performance of our modelling technique. The effect of pulse power and risetime on the resulting scattering is examined, along with the effects of γ_a, an intrinsic fibre parameter related to the linewidth of the Brillouin spectrum. The spatial and temporal evolution of the acoustic field is illustrated; the effect of the steady state value of this field on the 3WI is investigated. The steady state acoustic field strength is related to the extinction ratio of the pulsed source, and it is found that this parameter has a significant influence on the scattering. This type of modelling provides a rapid means of investigating SBS as a tool in fiber optic sensing.

A novel design of photonic crystal fiber (PCF) for broadband dispersion compensation is proposed. By changing the radius of the air holes placed in close proximity of the guiding core, the PCF fiber can realize effective broadband dispersion compensation. Through the proper design, the PCF fiber is shown to provide large normal dispersion up to several hundred ps/nm/km. Two examples of the PCF fibers for dispersion compensation of the standard single mode fiber and "True wave-RS" fiber are given. For example, by decreasing the diameter of second ring of air holes, the proposed PCF fiber is shown to provide large normal dispersion (up to -811 ps/nm/km), nearly ten times of conventional dispersion compensating fibers, and compensate conventional single-mode fibers within ±0.05 ps/nm/km over a wavelength range of the conventional C-band.

Raman fiber amplifiers (RFA) have been used in optical transmission communication systems in recent years due to their advantages over Erbium-doped fiber amplifiers (EDFA). Recently, the analysis of RFAs dynamic response and transient’s control has become important in order to predict system response to channel add/drop(s) or cable cuts in optical systems, and to avoid impairments caused by the power transients. Fast signal power transients in the surviving channels are caused by the cross-gain saturation effect in RFA, and the slope of the gain saturation characteristics determines the steady-state surviving channel power excursion. We are presenting the modeling and analysis of power transients control using two different approaches: (1) Power transient control using a Pump control method for a single and multi-pump scheme, and (2) the gain clamping of RFAs. Both methods are analyzed in a single amplifier as well in a cascade of RFAs.

Fiber Bragg gratings (FBGs) have emerged as important components and received intensive research attention in both fiber telecommunication and sensing fields. Bragg gratings in single mode fiber structure (SMFBGs) have been studied extensively. On the other hand, fewer studies have been reported on multimode fiber Bragg gratings (MMFBGs) despite of their potential applications in future optical access networks. In this paper, MMFBGs are studied in detail both theoretically and experimentally. A comprehensive numerical model is developed for MMFBGs based on the coupled mode theory and applied to analyze measured transmission and reflection spectra from MMFBGs with reflectivities ranging from 78% to 99%. It is found that the spectra of MMFBGs depend strongly on fabrication conditions (e.g. modulation depth of the grating) and experimental conditions (e.g. mode excitation). Good agreement is obtained between the theoretical simulations and experimental measurements. Our simulations based on the developed MMFGB numerical model can provide quantitative explanations for the observed experimental phenomena. These explanations give a complete understanding of the nature of the interaction between the wave propagation and multimode fiber gratings. It is considered that the spectral simulations provide a theoretical guidance in design MMFBG based devices.

Photonic crystal fiber (PCF) is a new type of optical fiber that adds a unique dimension to fiber design possibilities, enabling the previously unthinkable in fiber optics. The mode and the band-diagram analysis by conventional electromagnetic methods had proven the advantages of the PCF. When this new product couples to the existing optical fiber systems or vice versa, the important issue is the coupling efficiency. From among the available numerical techniques, the FDTD method seems to be the most reliable for such a task. The method does not force any assumptions on the behavior of the propagating light. It solves Maxwell’s equation in time domain and space domain directly and is readily applicable to problems like transmissions through refractive index discontinuities and multiple reflections on interfaces. The paper will present FDTD simulation results of the coupler and analysis of the calculated coupling efficiency on coupler design parameters.

The increasing demand for high capacity optical networks and the decreasing revenues per bit, combined with the given economy of scale for optical networks, forces the network operators to enhance the channel data rates as well as the channel numbers. Higher channel data rates result in a lower footprint, energy consumption and a lower complexity in network management and operation support systems, due to lower channel numbers. The enhancement of channel data rate in principle leads to a system tolerance reduction for chromatic dispersion, PMD and nonlinear effects. Furthermore higher order effects like dispersion slope and higher order polarization mode dispersion have to be taken into account. On the other hand the fast pulse broadening leads to a quasi linear behaviour of the systems, which relaxed some link design rules compared to 40 Gbit/s transmission. The lower tolerances can partially be mitigated by the implementation of more complex amplification schemes and compensators. The complexity of system design, accounting for less tolerances and adaptive compensating modules, is increased. We investigate theoretically and numerically the limiting physical effects and the impact on the signal performance, induced by chromatic dispersion, PMD and nonlinear impairments. We present derived engineering rules for all relevant effects and for various fiber types, based on channel data rates of 160 Gbit/s. These engineering rules enable design engineers to perform a fast system design and system degradation estimation, without time consuming full numerical simulations.

Chromatic dispersion accumulated over a link of optical fiber causes transmission degradation for data rates of 10 Gbit/s and higher. The different technologies for compensating this effect are reviewed. They are compared based on their suitability for different applications. Characteristics such as system performance, tunability, insertion loss, bandwidth and size are discussed. Tunability is of increasing importance as the future networks become more and more agile and reconfigurable. Fine tuning is also a requirement at data rates of 40 Gbit/s and higher. In the context of WDM systems, operations in both single-channel and multi-channel regimes are discussed. Recent developments are presented and future trends are discussed.

For the last decade, recirculating loops have been a useful tool in the research and development of long haul transmission links. A loop experiment can emulate the transmission of an optical signal over thousands of kilometers by using a relatively short link of a few hundred kilometers and recirculating the signal several times. Although recirculating loops accurately replicate most physical effects encountered in point-to-point links (loss, noise, chromatic dispersion, nonlinear effects, etc), the statistics of polarization effects (polarization mode dispersion (PMD) and polarization-dependent loss (PDL)) may not be properly emulated. In an optical link, PDL can induce statistical fluctuations of the optical signal-to-noise ratio (OSNR) and consequently of the bit-error-rate (BER). Due to environmental changes, the effects of PDL vary stochastically in time. The periodic nature of fiber loop may artificially produce an unrealistic PDL distribution and the statistical distribution of PDL effect may be significantly different from that in a installed link. We report the analysis and observation of a power oscillation effect caused by PDL due to the periodic nature of the polarization evolution in a recirculating loop. The oscillation is expected to affect the OSNR and consequently the BER as a function of recirculation.

Polarization-mode-dispersion (PMD) tolerance of a duobinary modulation format is investigated. With the use of 2.5 Gb/s Mack Zehnder modulator for modulation and optical pulse shaping of a 64 channels (10.7 Gb/s) dense wavelength division multiplexing metro networks, the electrical filter needed in conventional duobinary is no longer needed. The proposed method cut down in the complexity and cost of the duobinary transceiver. This technique allows the use of 10 Gb/s MZ modulator for 40 Gb/s duobinary transmission. We also used electronic compensation to adaptively compensate for PMD/GVD.

The effect of input pulse chirp on the optical eye diagram for communication systems having highly mode coupled PMD, PDL and CD is reported via analytical evaluation. It is shown zero chirped input pulse is not always the best choice. As well, we study the effect of duty cycle of the input pulse on the optical eye diagram for communication systems.

Multicanonical Monte Carlo simulations are used to investigate the joint probability of polarization mode dispersion (PMD) and polarization dependent loss (PDL) in a field fiber emulator. A joint PDF of PMD and PDL showing rare events is reported for the first-time.

Distributed Raman amplifiers (DRAs) are an enabling technology for long- haul and metropolitan- area broadband optical networks. These devices utilize the stimulated Raman scattering process in order to achieve gain over the bandwidth of ~ 40 THz from the transmission fiber itself. While DRAs’ offer various advantages by their ubiquitous presence in the transmission path, they also pose challenges such as gain flattening in the broadband spectral regime. Evidently, one cannot use fiber Bragg gratings (FBGs) and other filtering devices such as thin-films filters. However, gain flattening over wideband can be achieved using multi–wavelength (multi-λ) pumping [1], that is accomplished with spectral slicing based on Raman fiber lasers (RFLs). Novel test amplifiers have been designed and simulated with various pump parameters such as the number of RFLs pumps, their wavelengths, and relative powers. The results obtained using a 6-λ pump module in the range 1460-1510 nm, showed a better gain-flattened amplification [2]. We have designed various network test topologies, which are simulated using commercially available software packages. A 6- λ pump module using 2 to 5 pairs of fiber Bragg gratings [2] seems optimal. The output FBGs are tunable in order to provide reconfigurable pump module. Details of the results and design optimization will be presented. [1] Y. Cao, M. Y. A. Raja, “Gain-Flattened Ultra Wide Band Fiber Amplifiers,” J. Opt. Eng., Vol. 42, No. 12, 3347-3451 (2003). [2] Y. Cao, J. G. Naeini, K. Ahmad, and M. Y. A. Raja, “Gain-flattened Distributed Raman Amplification using Multi-l Raman Fiber Laser Pumping”, presented in OISE’03, Orlando, FL.

Reels of fiber are generally used to test fiberoptic components and compare various network archetectures. It is possible in this fashion to measure the bit-error rate, power penalty and other characteristics of a nominal fiberoptic link within the confines of a laboratory. However such measurements cannot easily be automated and practical considerations tend to limit the number of test cases. As a consequence stress testing, or pushing test conditions beyond the nominal case as well as the systematic measurement of design margins has not been widely available to designers of fiberoptic components and networks. Optical fiber emulators which replace reels of fiber with advanced electro-optical modules to produce a controllable amount of signal degradation address many of these issues. The design of such emulators presents many challenges which are discussed in this paper along with specific test cases.

As optical communication systems become more complex the quality of signals can be significantly affected by polarization mode dispersion (PMD) effects from optical fiber and in-line components. While the PMD is a vector quantity with a magnitude differential group delay (DGD) and a direction principal state of polarization (PSP), the interest was focus on the DGD value. Vendors demand from optical components manufacturers that the DGD introduced by a devices to be below a certain value. For this reason it is imperative to be able to accurately measure the PMD effects. In this report we present our investigation with respect to the accuracy issues related to the three techniques used for the PMD measurement. In each case the advantages and the drawbacks are presented. We have selected these methods because they are three among the four methods suggested by ITU-T under Recommendation G.650(1997 modified 2000).

Passive optical interconnection is being used in large packet switches to allow size scale up and more efficient heat dissipation from the electronic processors. The line card is the electronic island whose size is determined by thermal dissipation technology. We have selected an optical interconnection technology that is low cost, fiber ribbon, and we are working on the shutter arrays, which allow active routing. This will reduce the demands on the electronic processors. Progress on a 180 Gbps switch will be reported.

In WDM optical networks, signals have to be manipulated and monitored on a per wavelength basis. This requires that (de)multiplexing and wavelength processing functions are combined in the same optical component. The most efficient solution from a footprint size, cost and reliability point of view would be a combination of both functions in one photonic integrated circuit (PIC), fabricated by photolithography. Given that processing / detecting of individual wavelengths is an active function, in S, C and L communication bands naturally realized by InP and related semiconductor materials, InP-based PICs are widely agreed to be the superior choice. This paper reports a practical design of such InP-based PICs. It is based on a building block approach, which allows a large variety of WDM optical components to be built from relatively few monolithically integrable elements, by using standard semiconductor fabrication technologies. These include: (i) (de)multiplexer based on a planar echelle diffractive grating with polarization compensation in the slab waveguide region; (ii) single-mode vertically integrated waveguide active devices with detecting, attenuating, amplifying and switching features, inserted in every (in)output waveguide channel of a planar (de)multiplexer, and (iii) passive waveguide circuitry required for coupling the light to and directing it through the InP-based photonic chip. Design and characterization examples of the building blocks and PICs for channel monitoring, variable attenuation / equalization and spatial switching are presented.

Agile optical amplifiers must preserve constant gain and flatness over a wide range of input conditions to meet the needs of reconfigurable optical networks. Though routers and switches used to perform system protection or network reconfiguration feature slow transition times, reported theoretical and experimental analyses of the dynamics of agile EDFAs have focused on fast, sub-microsecond input-signal transitions. As a result, only low population inversion, one- or two-stage baseline amplifiers practically provide acceptable transient performance over the entire dynamic and wavelength range of operation. On the other hand, advanced, versatile optical amplifiers, which are highly desirable to perform system-level controls such as gain adjustment or reconfigurable wavelength add-drops at mid-stage, feature large excursions in these input conditions. In this paper, we discuss the line-transient times in view of today's switching technologies and Standards and provide experimental evidence of the suitable performance of advanced agile amplifiers at system switch speeds. The amplifier used is a very high-performance, agile device, which relies on a 3 μs feedback loop. Input transients up to 15 dB were studied, showing excursions smaller than 1 dB and steady-state gain errors no larger than 0.2 dB.

This paper presents a thermo-mechanical analysis of an optoelectronic system including a Mach-Zehnder optical modulator integrated with a broad-band GaAs driver amplifier, forming a module which then is placed into a low temperature co-fired ceramic (LTCC) substrate. All module connections such as voltage supply, RF signals and fiber optic input/output are realized through the LTCC. Thermal analysis of this integrated system shows elevated temperatures in the optical component caused by the heat generated in the power amplifier and dissipated into the substrate-carrier and from there into the LTCC. Temperature profiles along the MZ modulator reveal a strong non-uniformity, reaching a 26C temperature difference between the optical component input and output. A stress-strain analysis is also performed. Preliminary results show significant physical distortion of the optical component, which could cause optical misalignments and additional coupling losses. These findings indicate a need for thermal consideration in early design stages.

In this paper, a three-stage packet switch architecture is implemented consisting of a reconfigurable optical center stage surrounded by two electronic buffering stages grouped into sectors to ease contention. A Flexible Bandwidth Provision algorithm is used to change the configuration of the optical center stage to form the requested bandwidth desired by incoming traffic. The switch is modeled by a bipartite graph built from the service matrix. The bipartite graph is decomposed by solving an edge-coloring problem and the resulting permutations are used to configure the central stage removing the requirement for a per-time slot scheduler. Flexible Bandwidth Provision (FBP) algorithm requires dynamically reconfigurable technology readily available in programmable logic devices. The designed packet switch being a collection of discrete entities is most easily implemented on separate programmable logic devices forming electronic “islands” interconnected by photonics technology. The demonstrator itself contains 64 inputs and 64 outputs with reconfigurable central stage crossbars. The switch is a collection of input and output sectors each implemented on a single FPGA. Each sector is an 8 x 8 sub-switch with shared buffer memory. The interface between the sectors and the central stage will use VCSEL technology for O-E-O conversion. The input sectors together with the central stage form the adaptive portion of the switch configured by an embedded soft-core processor implementing the FBP algorithm of which is entity are located on an Ethernet local area network. This switching architecture has also been simulated and results show that this architecture result in a dramatic reduction of complexity, at the price of only a modest spatial speed-up (<2).

Communication transmission systems continue to evolve towards higher data rates, increased wavelength densities, longer transmission distances and more intelligence. Further development of dense wavelength division multiplexing (DWDM) and all-optical networks (AON) will demand ever-tighter monitoring to assure quality of service (QoS). Traditional monitoring methods prove to be insufficient. Higher degree of self-control, intelligence and optimization for functions within next generation networks require new monitoring schemes to be developed and deployed. Both perspective and challenges of performance monitoring, its techniques, requirements and drivers are discussed. It is pointed out that optical layer monitoring is a key enabler for self-control of next generation networks. Aside from its real-time feedback and safeguarding of neighboring channels, optical performance monitoring ensures the ability to build and control complex network topologies while maintaining an efficiently high QoS. Within an all-optical network testbed environment, key performance monitoring parameters are identified, assessed through real-time proof-of-concept, and proposed for network applications in safeguarding of neighboring channels in WDM systems.

The dominant limiting factors affecting the performance of an amplified 10-Gbps CWDM data transmission system with an inline semiconductor optical amplifier (SOA) are investigated. More than 3 dB of system penalty at a BER of 1x10^-9 can be attributed to ASE and cross-gain-modulation (XGM) effects.

For a source-destination pair to communicate in a connection-oriented wavelength-routed optical network, a connection in the optical layer between the two nodes must be established. This process, also known as Routing and Wavelength Assignment (RWS), is realized by selecting a path between the two end nodes and allocating a suitable wavelength. The aim of the RWA process is to find routes and assign wavelengths for connection requests in a way that minimizes the consumption of network resources, while at the same time ensuring that no two lightpaths are assigned the same wavelength on a shared fiber link. Routing and wavelength assignment in wavelength-routed WDM networks is a major design issue, especially when survivability is a requirement. To minimize resources in such networks operating under static traffic environment, the problems of routing and wavelength assignment must be solved jointly as a single problem. This study proposes a different approach to formulate the problems of routing and wavelength assignment as Integer Linear Programming (ILP) problem. Unlike other formulations, where the routing sub-problem and wavelength assignment sub-problem are considered separately, this approach addresses the RWA problem compounded. Although this approach increases the number of variables in the problem, it guarantees the optimal solution. Furthermore, the problem may in many cases be solved using simpler linear programming techniques.

Explosion of data and mass availability of Internet connections around the globe had created huge bandwidth requirements for bandwidth hungry applications. Despite the technological advances in the core and their ability to transport, still much work has to be done in the access networks in order to be able to let the broadband traffic be transmitted transparently. Current solutions for access networks do not provide a concrete solution for the famous last mile problem. In this review we will investigate optical access networks as viable solutions to the ongoing problems in the access networks. Further we will study physical and technological limitations with the current state-of-the-art optical technology. We will address the optical access networks promises in responding to these shortcomings. We will thoroughly study passive optical networks (PONs), reviewing different type of PONs considering their benefits and limitations. We will conclude our study with comparison of these current solutions.

Evolving carrier networks are starting to include optical networking, but current optical elements form islands of communications, not the utopian “global intelligent optical network” that is the end goal. Although dense wavelength division multiplexer (DWDM) technology leverages the cost and availability of fiber, each wavelength in DWDM links needs its own transponder, an expensive O-E-O (opto-electrical-opto) device in which optical signals are converted back to their electrical (binary) format to process the information. In today’s carrier networks, O-E-O is needed to terminate each optical end point in order to deliver it in its correct format to the specific addressees using high-level format-dependent protocols. The presentation will discuss how 60-80% of all traffic that passes through an O-E-O switch at midpoints in a network typically does not require processing at those points and, in fact, get re-converted and re-transmitted. The equipment that does this costs millions of dollars; takes up more space; uses a lot of power (expense and heat); and creates additional opportunities for network failures. Adding O-O-O (photonic, or all–optical) switches to the architecture as needed allows higher speeds, less real estate, lower capital and operational expenditures, improved network reliability, and better customer response.

Distribution of millimeter-wave signals over optical fiber has been considered a promising technology for future broadband wireless access networks, thanks to the low loss and broad bandwidth of optical fibers operating at the 1550 nm window. Different schemes have been proposed to distribute millimeter-wave signals using optical fiber, which include intensity modulation and direct detection (IM/DD) scheme and remote heterodyne (RHD) scheme. In a millimeter-wave-over-fiber system using IM/DD scheme, two sidebands located at the two sides of the optical carrier are generated. For frequencies higher than 20 GHz, the chromatic dispersion becomes a serious problem which leads to high power penalty. The dispersion problem can be solved if RHD scheme is used. In an RHD scheme, two wavelengths that are phase correlated are generated using single-side band with carrier modulation, optical carrier-suppressed modulation, optical offset injection locking or optical offset phase locking of two laser sources. Ideally, the laser sources are considered to have very narrow linewidth, which will not introduce phase noise at the remote side when beating the two wavelengths. However, in real applications laser diodes usually have a finite linewidth, which leads to the phase de-correlation in the fiber links; phase noise is then generated at the remote end. In this paper, we will analyze the effects of the finite linewidth of optical sources on the performance of millimeter-wave over fiber systems. Simulation and experimental results will be provided.

Radio signal transmission over optical fiber is considered a promising technology for future wireless communication applications, thanks to the advantages such as broad bandwidth, low loss, and the immunity to the electromagnetic interference. There are two schemes that are employed to transmit radio signals over optical fiber, the intensity modulation/direct detection (IM/DD) and remote heterodyne detection (RHD). For IM/DD scheme, the RF signal is intensity modulated onto the optical carrier at the transmitter and recovered by a photodiode at the remote end. The problem related to IM/DD scheme is that the chromatic dispersion caused by optical fiber leads to significant power penalty. To avoid chromatic dispersion-induced power penalty, RHD scheme is usually used. For RHD scheme, the RF signal is carried by two phase-locked optical carriers and is recovered at the remote end by a photodetector. On the other hand, there have been great interests in the development of true time delay (TTD) beamforming techniques for phased array antennas. Phased array antennas using TTD techniques can provide wide instantaneous bandwidth without beam squint problem, which ensures an accurate delivery of radio signals. Different TTD beamforming techniques have been proposed, but all of them are based on IM/DD scheme, which has the problem of power penalty. In this paper, we propose to use RHD scheme for TTD beamforming, to avoid dispersion-induced power penalty. A fiber Bragg grating prism for broadband beamforming using discrete fiber Bragg gratings will be designed and implemented. Theoretical and experimental results will be reported.

Switching is one of the key functionalities in next generation optical networks. It might be performed by either an optical switch (optical-electrical-optical, or OEO) or a "purely" photonic switch (optical-optical-optical or OOO). Both switches are analyzed from two perspectives - as an individual network element, and as an integral part within the communication network. As an individual network element, the performance evaluation of the two switch types is based on the individual assessment of switch footprint and power dissipation, bandwidth utilization, scalability to high speed, transparency, interoperability, technology maturity and ability to manipulate data. Although both switch types have their own advantages as a network element, the full judgement of their role in next generation optical networks requires an overall network perspective. From that viewpoint, network functionalities such as grooming capabilities, scalability, traffic management, protection, line equalization and performance monitoring are those taken into account for comparative analyses to gain an understanding of the impacts of switch choice in the network. As a result of the comparative performance assessment, the merits and benefits of both switch types in actual network applications are analyzed and outlined. Although the paper evaluates some criteria for switch choice in a network, it points out potential technologies or techniques critical to next generation architectural solutions and protocols as well as the challenges to bridge the gap towards implementing flexible, cost-effective and dynamically provisioned networks of the future. Finally, the paper responds to one critical question - What is the expected role of each switch type in next generation applications and services?

Recently pilot tones have been widely deployed as a path supervisory method for optical crossconnects (OXCs). In this work we present a wavelength-routing fault detection scheme for concatenated OXCs in an all-optical network (AON) testbed, in which pilot tones are added to wavelength channels as channel identifiers (CIDs) at input ports. OXC routing errors then can be detected by comparing the CIDs at output ports with the stored routing information. The AON testbed is based on commercially available photonic switches, which support dynamic wavelength switching. At each input port of an OXC, a unique frequency tone is added. We compare the performance of two sets of candidacy pilot frequencies, 101 kHz ~ 117 kHz and 1.01 MHz ~ 1.17 MHz (with 2 kHz and 20 kHz separation respectively). The modulation index is set to 10%. On the output side of each OXC, a modulator is inserted after each output port. We detect the tone after the modulator and feed the amplified, filtered, and inverted signal back to the modulator, for removing the tone. The pilot tones added to all OXCs construct the concatenated wavelength-routing fault detection scheme. This work numerically evaluates the effects of concatenated pilot tones and different pilot frequencies on the overall system performance, e.g., bit error rate or Q-factor. The simulation results show that the proposed scheme is feasible and the degradation of system performance due to pilot tones is negligible.

Recently, people pay more attention to the grid technology. It can incorporate high-speed network, high performance computer, large database, sensor and remote equipment etc. Moreover, it can not only connect all kinds of resources in the network and also put them into a super transparent computing environment for customers. In which it can realize meta-computing for the shared resources. However, traditional parallel computing system such as SMP and MPP use multi-processors to upgrade computing by closed coupler. Doing like that, the flexible and scalable performance of the system will be limited so that the system can't meet the requirement of the grid technology. In this paper, the architecture of cluster computing system applying in grid nodes is introduced. It mainly includes the following aspects. First, the network architecture is analyzed and designed. Second, it forms virtual computers with distributing computing including divided computing and shared computing. Last, it discusses the communication problems between the grid nodes. In other words, it discusses the realization of single map in order to meet all the requirements of customers sent to the grid nodes.

In the NGI featuring grid technology, as a foundation of grid Meta-Computing grid node must have large capability of data storage and processing. In this paper, parallel processing and computing environment for Multi-CPU and Multi-MEM is researched, the main part as follows. Part 1.optical routing-switching system design a. Adopting CWDM to realize interconnection for parallel high capability computers whose parallel scale are less than 16. The advantages are network simplification, lessening cross talk, reducing cost and have nothing to do with single computer. b. We take IP packet format to be data transmission format to match today’s Internet, being efficient and making it compatible with each other. c. Researching Odd-Even merging network to build OXC routing system. We have done the optimizing analysis of the number of optical switches needed in different parallel systems, designing routing control system featuring centralization and inside control. Part 2.optical routing-switching protocol design and simulation This paper designs switching protocol of Multi-CPU and Multi-MEM parallel processing system. The protocol can control different wavelength signals to route and switch effectively. Simulation shows that it can reduce probability of block and reach the requirement of data transmission.

Progress in optical switching technology currently faces several major obstacles. One of these is high power consumption, which quickly multiplies in cascaded switch configurations. As well, many implementations have long switching speeds and large footprints. An improved compact 1x2 digital optical switch (DOS) in InGaAsP/InP is presented, with experimental results compared to numerical modeling. The Y-junction waveguide switch operation is based on reconfiguration of output waveguide arms by carrier injection at the electrodes. We present experimental results of DOS fabricated with InGaAsP cores having bandgaps of 1.2um, 1.3um, and 1.4um. The results are compared with calculations of refractive index change versus carrier concentration in the different InGaAsP alloys. Additionally, wafer layer structure and waveguide parameters were redesigned to decrease, respectively, power dissipation and optical confinement. Switching current is significantly reduced from greater than 100 mA to about 20 mA, which not only provides power savings, but also results in less thermal overshoot in the switched optical pulse. The DOS has a measured switching contrast ratio of better than 12.5 dB, with a transition time of less than 5ns. Polarization dependence of switching contrast is also explored.

AOM-based pulse shaping permits precise spectrum slicing, which is useful for DWDM-based architectures. In essence, this method uses microsecond-duration radiofrequency pulses to completely control the spectra of femtosecond laser pulses, hence achieves dramatic temporal data compression. A multiple color sources from a mode-locked laser simplify the system design and operation. We demonstrated the AOM pulse shaper as the modulator and a CCD camera with 256 pixels as the receiver. The spectrum of a 200 fs Erbium Doped Fiber Laser (EDFL) pulse was dispersed across the AOM’s aperture (FWHM was 35 nm) and was then modulated in a conventional pulse shaper. We tested 87 channels with channel-spacing of 0.41 nm using a 518-MHz modulator, and 120 channels with channel-spacing of 0.29 nm using a 148-MHz modulator (in each case a 0.1 nm guard band was used). Starting from the original pulses, this modulation creates time slots of 43 ps and 63.4 ps respectively. The equivalent speed of the transmission will be 2.0 Tb/s and 1.9 Tb/s in a highly multiplexed system. The spectral efficiencies achieved in this experiment were ~46%, approaching the theoretical limit of 50% for On-Off Keying (OOK) modulation. A benchmark image was successfully sent over the test bed.

Cascaded second-order nonlinear interaction (&chi⁽²⁾);based wavelength conversion technique has attracted much attention due to its unique characteristics such as low noise and broadband, which are critical in fiber communication networks. In this report, wavelength conversions based on the newly proposed SFG-DFG (sum frequency generation - difference frequency generation) and conventional SHG-DFG (second harmonic generation - difference frequency generation) are studied and compared both experimentally and theoretically in a LiNbO₃ quasi-phase matched (QPM) waveguide. It is shown that the same conversion efficiency can be achieved by employing two pump sources with only half power each (P1, P2) in the SFG-DFG scheme as compared with the SHG-DFG scheme with a single higher power pump beam (P=P1+P2). It is shown that the cascaded SFG-DFG based wavelength conversion has a larger 3-dB pump tolerance bandwidth. The theoretical results are consistent well with the experimental ones. It is found that the pump wavelength difference can be separated by a span as large as 75 nm, while 3-dB signal conversion efficiency is retained in a 45 mm-long device. It is also exhibited that tolerance of temperature for the cascaded SFG-DFG remains the same as that of the cascaded SHG-DFG based devices. The results show that the SFG-DFG wavelength conversion scheme is very attractive for practical applications.

A novel polarization scrambling optical signal-to-noise ratio (OSNR) monitor has been proposed and demonstrated. The OSNR monitor consists of a polarization scrambler, a polarizer and a photodetector. When the channel signal input to the polarizer is a linear polarization state aligned parallel (orthogonal) to the polarizer the output reaches its maximum (minimum). The OSNR can be obtained from the measured maximum and minimum of the output power if the sampling time is long enough to ensure a good coverage on Poincare sphere. The new OSNR monitor is polarization mode dispersion insensitive. The effect of polarizer extinction ratio and sampling time on the accuracy has been discussed.

The effective bandwidth of a wireless communication system is proportional to the carrier frequency, shifting the operating frequency of the wireless system from the crowded microwave L and S bands to the unregulated mm-wave band is a trend for future broadband wireless services. Intensity modulation and direct detection scheme (IM/DD) has been considered a simple method to impose a millimeter-wave signal onto optical fiber. However, for systems using IM/DD, the chromatic dispersion introduce significant power penalty, which limits the transmission distance. Remote heterodyne schemes have been proposed to solve the dispersion problem. Several approaches have been proposed to generate two phase-correlated optical wavelengths that are separated at a required millimeter-wave frequency. These approaches include single-side band with carrier modulation, optical carrier-suppressed modulation, optical offset injection locking and optical offset phase locking of two laser sources. All these methods provides phase-correlated optical wavelengths, but with complicated system configuration and high system cost. In this paper, we propose a simple method to generate phase-correlated wavelengths using a phase modulator and two fiber Bragg gratings. Two wavelengths are generated by modulating the phase modulator with an RF frequency. The required millimeter-wave signal is obtained by selecting two sidebands using the two narrowband fiber Bragg gratings. Theoretical analysis and experimental results will be reported in the paper.

In this paper,architecture of grid audio/video communication environment based on optical layer multicast is introduced. First, architecture of optical layer multicast communication environment is analyzed. Then, scheme of large capability optical routing is presented. Last, condition of non-blocking multicast communication over WDM(wavelength division multiplexing) grid network is analyzed.

Femtosecond laser ablation is an important process in the micromachining and nanomachining of microelectronic, optoelectronic, biophotonic and MEMS components. The process of laser ablation of silicon is being studied on an atomic level using molecular dynamics (MD) simulations. We investigate ablation thresholds for Gaussian laser pulses of 800 nm wavelength, in the range of a few hundred femtoseconds in duration. Absorption occurs into a hot electron bath which then transfers energy into the crystal lattice. The simulation box is a narrow column approximately 6 nm x 6 nm x 80 nm with periodic boundaries in the x and y transverse directions and a 1-D heat flow model at the bottom coupled to a heat bath to simulate an infinite bulk medium corresponding to the solid bulk material. A modified Stillinger-Weber potential is used to model the silicon atoms. The calculated thresholds are compared to various reported experimental values for the ablation threshold of silicon. We provide an overview of the code and discuss the simulation techniques used.

A wide band FM light source is useful for many applications in the areas of communication, spectroscopy and sensing. Chirped light can be generated by modulating the laser output externally with a phase modulator. But the spectrum is usually not wide enough due to the modest modulation depth of available commercial modulator. A more efficient method is to perform phase modulation of light inside the laser cavity, which under certain cavity detuning conditions can perform deeper FM modulation through cavity mode coupling effect. The output spectrum is much wider than that obtained external modulation alone. In this paper, we show the theoretical and experimental result of a 40GHz FM oscillation fiber laser. The detuning frequency and single-pass modulation depth can affect the final spectrum width. The wide spectrum of 1.42THz around 1560nm is obtained in the experiment. By using single mode fiber as dispersion medium only, we can compress the constant intensity light with 1.42THz bandwidth to 821fs pulse. This laser appears to be of great interest for study of high-speed optical communication systems for the impairments and their impacts at wide frequency range. Or, its ultra-wide, chirp output with high sweep rate will be useful for rapid and extremely sensitive interferometric measurements.

We have experimentally demonstrated ultrashort optical pulse generation from a continuous wave (CW) laser using an external electrooptic deflector (EOD). Highly efficient EOD operating at 16.25 GHz has been realized with periodically domain-inverted LiTaO₃ crystal. The shape of domain inversion region has been theoretically designed in consideration of the velocity mismatching between the modulation microwave and the light so as to realize spatially linearly-varying phase shift. The deflected CW Ar laser beam passed the Fourier transform lens, and then optical pulses were picked out at a repetition rate of 32.5 GHz through a narrow slit, which was placed at the focal point of the lens. The pulses were observed by streak camera (Hamamatsu: C5948) and the achieved shortest pulse width was estimated to be 0.9 ± 0.1 ps.

We have investigated the full three dimensional momentum correlation between the electrons emitted from strong field double ionization of neon when the re-collision energy of the first electron is on the order of the ionization potential of the singly charged neon ion. We find that the momentum correlation in the direction perpendicular to the laser field depends on the time difference of the two electrons leaving the ion. Our results are consistent with double ionization proceeding through transient double excited states that field ionize.

Past works have used high-order harmonics in gas targets to demonstrate attosecond pulse generation. However, recent theoretical simulations have shown that solid-surface harmonics can also be used to produce attosecond pulses. Solid-surface harmonics are generated when a high-intensity femtosecond laser pulse irradiates a solid target surface at an oblique incidence angle. The conversion efficiency of this phenomenon increases rapidly with increasing pump laser intensity, and there is also no presently known upper limit in the pump intensity that can be used. Accordingly, this method possesses the potential for high-energy attosecond pulse generation. The main aim of this paper is to experimentally clarify the optimal conditions for highly efficient solid-surface harmonic generation. We demonstrate up to the 16th harmonic (49.1 nm wavelength) of a Ti:sapphire laser using modest pump intensities of 4×10¹⁶ W cm^-2 irradiating a silicon wafer target. Investigations with low-order harmonics have revealed a large dependence of harmonic conversion efficiency on the target material. Furthermore, a drastic increase in the harmonic intensity has been observed by repetitively irradiating a metallic-coated target.

Ultrashort-pulse single-maximum nondiffracting beams of microscopic radius and large axial depths are interesting for applications in nonlinear optics and spectroscopy, for acceleration and manipulation of particles, measuring techniques, materials treatment or information processing. Here we report on the experimental generation of such beams by self-apodized truncation of Bessel and pseudo-Bessel beams from a Ti:sapphire oscillator. Small angle operation was enabled by thin-film structures. To obtain self-apodization, the diameter of the truncating diaphragm was adapted to the first minima of Bessel distribution. The propagation of (a) Bessel beams of meter-range axial extension shaped by axicon mirrors, and (b) microscopic pseudo-Bessel beams of millimeter-range extension shaped by Gaussian-shaped microaxicon lenses was studied. In case (a), single-maximum beams of > 20 cm depth were produced. To generate comparable focal zones from Gaussian beams, a much larger distance (10x) is necessary, and axial stretching of spectrum destructs the temporal structure. In case (b), the focal zone length was increased by a factor of >5 compared to a Gaussian beam. Arrays of truncated Bessel beams were generated as well. The experimental results indicate that truncated Bessel beams enable more compact setups than corresponding Gaussian beams and are in particular advantageous for ultrashort pulses. Further improvements are possible by combining coherent addition in resonators with truncation outcoupling.

Active FM harmonic mode-locking of a fiber laser by intracavity phase modulation allows obtaining of stable laser pulses with a high repetition rate. Driving the phase modulator by an external RF synthesizer has an advantage of constantly applying only one modulation frequency to the phase modulator. That greatly facilitates the generation of stable laser pulses with very small noise. However, this approach requires constant frequency tuning of the synthesizer to compensate for small temperature fluctuations causing changes in the fundamental frequency of the laser cavity. In order to control the modulation frequency we mixed the RF signal from the laser output detected by a fast photodiode with the signal from the synthesizer. The amplitude of the measured DC component of the mixed signal depends on the phase difference of the two signals. The phase difference varies approximately linear with the laser detuning near the mode locking resonance. We develop software that performs constant measurement of the mixed signal and tuning the modulation frequency in order to keep the DC component of the mixed signal at a preset value. The program performed approximately two auto-tuning steps per second. The presented method allows very simple and reliable obtaining of stable computer controlled harmonic mode-locking of a fiber laser at 40 GHz repetition rate frequency.

Pump-probe spectroscopy of molecular systems requires high average power, short pulse, mid-infrared sources. Today OPO deliver wavelengths up to 4 µm and THz systems supply wavelengths beyond 20 µm. To achieve tunable wavelengths in between these two regions, the signal and idler beams of the OPO can be difference frequency mixed again. This two-step nonlinear process necessarily leads to average power much less than the OPO pump, which is typically a 1 W Ti:sapphire laser. In this paper, we report initial experimental results of two-wavelength amplification in a Yb-doped DCF fiber. Yb:fibre amplifiers have been shown to deliver short pulses at average power levels of up to 20 W[1] .The goal of the work is to generate high average power (<1W) pump and signal beams for difference frequency generation at around 15 µm. A schematic of the experimental is shown in Fig.1. A photonics crystal fiber is pumped by a mode-locked Ti:Sapphire laser to create supercontinuum, from which we can select the two seed wavelengths by using a standard zero dispersion grating line that has been adapted to have individual mirrors in the focal plane for the two different colours. The Yb:fibre is pumped by a 915nm diode laser. As shown in Fig. 2, at a pump power of 2.1 W, 500 gain was achieved yielding 150 mW total power(1055 and 1095 nm). References:1. D. Nickel et al, Opt. Commun. 190, 309 (2001)

Terahertz (THz) radiation is ideal for probing many different materials and processes. Photons in the THz regime have energies on the order of an meV, which is an important energy scale for many electronic processes. In this paper we will describe the use of optical rectification of 50 fs IR pulses to generate THz pulses. Using this method, spectrally broad THz pulses with durations on the order of ps can be produced. This feature allows us to obtain time and frequency resolved information about the transmission of THz radiation during transient processes. A 50 fs IR pulse is used to optically excite a material and the relaxation as a function of time can be observed with the THz probe. We are developing a transient THz spectroscopy to study non-equilibrium processes in thin film superconducting YBa₂Cu₃O_7-δ (YBCO). We use ultrafast optical pulses to excite the sample, breaking a fraction of the Cooper pairs responsible for the film's superconductivity. This process produces highly energetic quasiparticles which thermalize and recombine on a picosecond timescale as the superconducting state recovers. Transient THz spectroscopy allows us to follow the evolution of this process with the required resolution, while simultaneously providing valuable spectroscopic information.