Silicon Photonics is emerging as an attractive technology in order to realize low cost, high density integrated optical circuits. Realizing active functionalities in Silicon waveguiding structures is being pursued rigorously. In particular, the Stimulated Raman scattering process has attracted considerably attention for achieving on-chip light generation, amplification and wavelength conversion. This paper reviews some of the recent efforts in using the Raman nonlinear process to realize amplifiers, and lasers. First the prospects of Raman process in realizing high gain amplifiers are discussed theoretically. Following this experimental results on amplification with gains as high as 20dB are presented. Some of the recent results in realizing pulsed and CW lasers with reverse-biased carrier sweep out are presented. The paper is concluded by highlighting some of the applications of the Raman process in Silicon in realizing mid-IR sources and also the use of SiGe as a flexible Raman medium are discussed.

Bismuth based erbium doped fiber and planar waveguide exhibit inherent features which cannot be realized with silica based fibers and waveguides. Extend L-band amplification, high gain C+L band amplification for coarse WDM and short pulse amplification without spectral broadening can be realized using bismuth oxide based EDF and 1-cm²-size Er doped spiral waveguide which shows >15 dB gain can be realized using bismuth oxide glass.

Semiconductor optical amplifiers (SOAs) based on nanostructure gain media such as quantum dots (QD) and quantum dashes (QDASH) have several basic characteristics which offer significant performance improvements over commonly used quantum well (QW) or bulk amplifiers. Among these are broadband optical gain bandwidth (which is two to three times broader than that of QW/bulk gain media), fast gain dynamics, large saturation powers, and low α parameter and population inversion factor. Originally, these properties have been demonstrated for QD/QDASH SOAs operating at 1000 nm and 1300 nm. However, it is imperative that QD/QDASH SOAs operating at 1550 nm be materialized in order for them to have the expected impact on fiber-optic communication. Operation at 1550 nm has been achieved using InAs / InP QD and QDASH laser structures. In this paper the unique gain and noise properties of InAs / InP QDASH SOAs operating at 1550 nm will be presented. Specifically, cross-gain-modulation, four-wave-mixing and chirp measurements which explore the complex spectral cross relaxation dynamics of these SOAs will be described and highlighted in the context of simultaneous, distortionless, high bit-rate multiwavelength data amplification, as well as wideband / high-speed optical signal processing applications. Also, an experimental study of the gain and noise in saturated QDASH SOAs will be described together with a theoretical analysis comprising both coherent and incoherent gain phenomena. The impact of the partially inhomogeneously broadened gain spectrum, fast population pulsation dynamics, α parameter and wetting layer density of states on the noise characteristics will be discussed.

Fiber optical parametric amplifiers (OPAs) are nonlinear devices based on the third-order nonlinear susceptibility of glass fibers. They require one or two pumps, located near the fiber zero-dispersion wavelength, and phase matching between the waves must occur. Injecting a signal at the input results in amplified signal and a new wavelength, the idler, emerging from the output. Important features have been demonstrated, including: 60 dB of cw gain; 400 nm gain bandwidth; tunable narrowband gain spectra; noise figure below 4 dB. The presence of the idler can be used for wavelength conversion. The spectrum of the idler is also inverted with respect to that of the signal; thus by placing an OPA in the middle of a fiber span one can realize mid-span spectral inversion (MSSI) which counteracts the effect of fiber dispersion and some nonlinear effects. By modulating the pump one modulates signal and idler at the output. This can be used to implement a variety of signal processing functions, including: demultiplexing of TDM signals; retiming and reshaping functions (2R regeneration). Some challenges must be overcome for fiber OPAs to be useful in communication applications. In WDM systems, these are: four-wave mixing and cross-phase modulation between signals; cross-gain modulation. Fiber OPAs also exhibit conversion of pump RIN and/or FM (used to suppress SBS) to signal and idler IM. Fiber OPAs will benefit from design and fabrication of novel fibers with high nonlinearity and improved dispersion properties. Novel families of fibers, such as holey fibers, are promising in this respect.

After a brief introduction of optical bistable operation of semiconductor active devices such as LDs and SOAs, recent progress in polarization bistable VCSELs and their applications for all-optical signal processing are presented. Applications include all-optical flip-flop operation with very low switching energy and high repetition rate, all-optical signal regeneration, and optical buffer memory.

A Wide wavelength tunable laser is needed for Wavelength Division Multiplexing (WDM) and Reconfigurable Optical Add/Drop Multiplexing (ROADM) networks, since it realizes flexible network, effectively employing wavelength resources, and inventory cost reduction. Several types currently exist, but they all are difficult to produce; that is, their mass producibility is not high and they have many components. In particular, monolithically integrated wavelength tunable lasers, such as DFB array, and SG(Sampled Grating)-DBR based structures, have been developed. While these lasers have good performance, they require complex InP growth steps and processing. The external cavity lasers also have good performance, but require precise manual assembly and have moving parts. We have proposed novel tunable laser consisting of silica waveguide ring resonator connected directly to semiconductor optical amplifier. This laser structure has several advantages, such as a simple laser structure suitable for mass-production and high reliability due to having a stable thermal optic phase shifter and no moving parts. This paper gives recent progress in waveguide ring resonator based tunable laser. Low loss and high performance silica waveguide ring resonator, which was suitable for tunable laser, was successfully fabricated using high index contrast SiON core. Double-ring resonators successfully attained 45-nm and 160-nm wavelength tuning operations, which was the largest wavelength tuning range in a tunable laser with no mechanical moving parts reported to date. Triple-ring resonator demonstrated stable full L-band tuning operations with 50-GHz wavelength spacing. We believe that silica waveguide ring resonator based tunable laser is very suitable for not only mass production, but also widely wavelength tuning and stable single mode operations.

Recently, Coarse Wavelength Division Multiplexing (CWDM) systems expand in local access and metropolitan area optical networks. In CWDM systems, uncooled and directly modulated DFB lasers are used, and the eight wavelengths spaced by 20nm from 1470 nm to 1610 nm are multiplexed. Then, in the eight wavelengths, each laser has been demanded to have similar characteristics such as threshold currents and slope efficiencies. However, among the eight wavelengths, L-band (1590 nm and 1610 nm) DFB lasers are inferior to S, C-band DFB lasers in performances such as threshold currents and slope efficiencies, so that the improvement of L-band laser performances have been demanded. The inferior of performances of L-band lasers is said to be mainly caused by increases of Auger recombination and Intervalence band absorption (IVBA). We designed the S-band, C-band and L-band lasers to improve the internal loss values and minimize the difference of laser performance among these bands. In L-band, we achieved the successful wide temperature range (-40 °C to 95 °C) operation of 300 μm long DFB lasers. They showed the threshold current of as low as 40 mA and the slope efficiency of as high as 0.27 W/A at 95 °C. As a result, regardless of the peak-wavelength difference, the CWDM DFB lasers show the similar performances and the excellent high temperature L-I characteristics. These results show that these lasers are suitable for using as the uncooled light source in CWDM applications.

We propose a novel InP-based traveling-wave electrode Mach-Zehnder modulator. It has an n-i-n isotype heterostructure to reduce both electrical signal loss and the optical loss caused by the p- type cladding layer. This device provides a large modulation bandwidth of more than 40 GHz. We have also developed a compact push-pull driven modulator module. We obtained error-free operation for a 40-Gbit/s NRZ signal in a push-pull configuration with a very low driving voltage of 1.3 V_pp. We also confirmed that the modulator has low chirp characteristics by demonstrating a 100-km SMF transmission with a penalty of less than 1.5 dB for a 10-Gbit/s NRZ signal.

The paper presents the development and optimization of ultra-broadband photodiodes (> 100 GHz) and balanced photodetectors (70 GHz). These developments are based on the waveguide-integrated photodiode type applying semi-insulating optical waveguide layers, grown on InP:Fe. The chip developments, characterizations in the frequency and time domain and the packaging of the chips into modules with fibre pigtail and 1 mm connector are described. The modules are tested at bit rates of 80, 100 and 160 Gbit/s, employing NRZ, RZ and DPSK modulation formats.

Broadband photodetector having detection capability ranging from visible to near infrared can be useful as a common detector for both data and transport applications. The capability of using single detector for all application in optical communication as the receiver, not only makes the total system cost lower, but it also makes easier the system vendors to reduce the inventory. We proposed detector having the detection capability ranging from 650 nm to 1750 nm, wavelengths that covers all optical communication wavelengths application. This invited paper has two-fold objectives: (a) provide a comprehensive overview of conventional photodetectors and their types, being used in today's optical communication and (b) introduce a development of multi-wavelength photo detector which authors pioneered. The features of proposed multiwavelength detector are simple structure, low-cost, high quantum efficiency, high sensitivity, and high speed.

Lincoln Laboratory has developed 32×32-pixel ladar focal planes comprising silicon Geiger-mode avalanche photodiodes and high-speed all-digital CMOS timing circuitry in each pixel. In Geiger-mode operation, the APD can detect as little as a single photon, producing a digital CMOS-compatible voltage pulse. This pulse is used to stop a high-speed counter in the pixel circuit, thus digitizing the time of arrival of the optical pulse. This "photon-to-digital conversion" simultaneously achieves single-photon sensitivity and 0.25-ns timing precision. We discuss the development of these focal planes and present imagery from ladar systems that use them.

A new low-loss high-index-contrast photonics platform has been developed for integrated optics and microwave photonics. The platform consists of a material system that has an index contrast that is adjustable from 0 to 25% and which is processed using conventional CMOS tools. The platform allows one to four orders of magnitude reduction in the size of optical components compared with conventional planar technologies. As an example, meter long path lengths occupy coils that are millimeters in diameter. Microwave photonic building blocks that are enabled include large bit count programmable delay lines for beam steering and shaping that fit in less than a square centimeter and which have delays controllable from 5 fsec to 10 nsec. Also enabled are arrays of high order tunable filters, a hundred micrometers in size, having linewidths ranging from tens of MHz to tens of GHz. These filters can be tuned over several hundred GHz, and when placed in Vernier architectures can be tuned across the C band (5 THz). An optical chip typically consists of dozens of optical elements. Each element is placed in its own micro-control loop that consists of a thin film heater for thermo-optic control and a thermistor for electronic feedback. The micro-control loops impart intelligence to the optical chip.

This paper reviews photonic crystal (PC) based demultiplexers, and briefly reports our latest experimental achievement in ultra-compact, power-efficient silicon photonic crystal waveguide (PCW) modulators. We review the modeling techniques for photonic crystal superprism devices, which utilize anomalous refraction on a photonic crystal surface for wavelength demultiplexing. The finite difference time domain method tends to be time consuming for the superprism devices as such devices demand fine spatial grids to resolve fine wavelength difference. Other theoretical methods suffer a variety of other drawbacks. A general, efficient PC refraction theory that can handle any surface orientation is needed for scientific research and device design. We review a rigorous PC refraction theory that we recently developed for these needs. Essentially, the refraction problem can be rigorously solved by computing the electromagnetic field in only a single cell on the surface. A new concept, surface-orientation-dependent eigenmode degeneracy, is introduced to explain certain subtle effect that occurs when the surface orientation undergoes a slight change. In addition, the transmission of a Gaussian beam or other realistic beam profiles is discussed. A complete theoretical framework of the photonic crystal refraction and transmission has thus been established. The theory has been applied to design a high channel-count dense WDM demultiplexer with 3dB or lower losses. Lastly, a silicon PCW Mach-Zehnder modulator with an 80-micron interaction length is reported. The slow group velocity in PCWs is exploited to enhance the modulation efficiency and reduce the peak drive current to 0.15mA at a modulation depth over 90%.

Large scale integration of optical components on a single ship mimics the performance and cost benefits for electronic VLSI, but requires the ability to make sharp bends in optical waveguides. Sharp bends may be achieved by using high refractive index contrast between core and cladding to tightly confine the optical field in the core. However, for tightly confined fields, the evanescent wave extends only submicrons into the cladding, so that coupling between waveguides by leaky wave tunneling in a directional coupler requires high resolution by expensive lithography. We use analysis and the matrix method in simulations to show that by increasing the refractive index between the waveguides in the directional coupler, lithographic resolution and hence cost are reduced.

We review the athermalization techniques that have been proposed for AWGs and discuss the advantages and disadvantages provided by each approach. We then describe our recent progress on the design and fabrication of a silica-based athermal AWG with a 1.5%-▵ waveguide and report its compactness and excellent optical characteristics including its extremely low insertion loss.

A fully optical PCB with transmitter/receiver system boards and optical bakcplane was prepared, which is board-to-board interconnection by an optical slot. We report a 10 Gb/s PRBS NRZ data transmission between transmitter system board and optical backplane embedded multimode polymeric waveguide arrays. The basic concept of the optical PCB is as follows; 1) Metal optical bench is integrated with optoelectronic devices, driver and receiver circuits, polymeric waveguide and access line PCB module. 2) Multimode polymeric waveguide inside an optical backplane, which is embedded into PCB, 3) Optical slot and plug for high-density (channel pitch : 500 um) board-to-board interconnection. The polymeric waveguide technology can be used for transmission of data between transmitter/receiver processing boards and backplane boards. The main components are low-loss tapered polymeric waveguides and a novel optical plug and slot for board-to-board interconnections, respectively. The transmitter/receiver processing boards are designed as plug types, and can be easily plugged-in and -out at an optical backplane board. The optical backplane boards are prepared by employing the lamination processes for conventional electrical PCBs. A practical optical backplane system was implemented with two processing boards and an optical backplane. As connection components between the transmitter/receiver processing boards and backplane board, optical slots made of a 90°-bending structure-embedded optical plug was used. A 10 Gb/s data link was successfully demonstrated. The bit error rate (BER) was determined and is 5.6×10^-9(@10Gb/s) and the BER of 8 Gb/s is < 10^-12.

A 400 Gbps backplane switch was developed with low-cost, small-size, 8-channels 10 Gbps/port optical I/O and a SiGe Bi-CMOS switch LSI on a 60x60-mm² BGA package. It indicates the applicability of backplane switch for high throughput backplane interconnections.

A passive optical network (PON) comes to front for cost reduction of FTTH systems. The PON shares the cost of fiber and an optical line termination (OLT) among several customers. The "triple-play services", broadband internet access, video and telephone on the PON, play the role of an engine for explosive growth of the FTTH. The triple-play services on the PON consist of passive devices, optical splitters and wavelength division multiplexing (WDM) modules, and active devices, an OLT and several optical network units (ONU). This paper reviews the technologies for low cost passive and active devices and describes our arrayed WDM filter module for video distribution and bi-directional transceiver device for OLT and ONU on the PON.

The potential of the optical circuit packaging technology is discussed. Special attention has been paid to introduction of "Optical wiring" into the Printed Wiring Board level ("last 1 meter area") to overcome conventional electrical copper-based bandwidth limitations. Optical Surface Mount Technology (O-SMT) can be one of possible solutions in this field is reviewed. High efficiency and alignment-free coupling between optical wirings and optical devices is a key. O-SMT requires a method to change the beam direction from the horizontal to the vertical and vice verse in order to couple between optical wirings in an OE-board and OE-devices mounted on the board. A novel method using an "Optical Pin" has been proposed and investigated. Furthermore, an optical coupling method using a Self-Written Waveguide called "Optical Solder" has been investigated. Several applications of self-written waveguides using a green-laser and a photo-mask are demonstrated.

A new fabrication technique is proposed for the fabrication of optical waveguides for use in optical communication modules with large-core optical fibers. The proposed technique employs a mixed photopolymerizable resin containing two kinds of photopolymerizable monomers that are different in terms of both refractive index and spectral sensitivity. Visible light is irradiated into the resin through an optical fiber in order by take advantage of the self-trapping effect to form the core portion. Only the lowest refractive index monomer is polymerized, with the reaction proceeding from the end of the fiber tip. After the irradiation is over, a concentration gradient is induced in the low refractive index monomer due to the selective area polymerization, which brings about a counter-diffusion phenomenon of the monomeric materials. Diffusion of the low refractive index monomer causes the high refractive index monomer to move out into the region surrounding the core portion. All of the residual monomers are subsequently cured by exposure to UV light. The region with decreased concentration of high refractive index monomer forms a cladding layer. The resultant refractive index profiles of the waveguides were experimentally observed to be "W-shaped". The measured propagation loss of a 700-μm-diameter waveguide was 1.7dB/cm at 0.68-μm wavelength. We are convinced that this technology could serve to automate optical fiber connection and packaging processes in the assembly of optical waveguide modules. This technology is especially useful in short-haul optical communication systems requiring a large-core optical fiber.

Polymeric optical waveguide devices connectable to plastic optical fibers (POFs) fabricated by thermal replication using thermo-plastic and thermo-curable resins are presented. Optical waveguides with large core sizes of 1000 μm were fabricated, and low propagation loss of ~0.2 dB/cm for thermo-plastic resin waveguide at 650 nm was achieved. Waveguide with high thermal resistance was realized by simultaneous curing and replication using thermo-curable multifunctional methacrylate monomers.

The stability of optical communications depends on the specification stability of optical components and on the properties of materials conjugated with the electromagnetic properties of light. Diffraction, the product of interference and the periodic variation of refractive index, is applicable in a plethora of optical devices that are deployed in dense wavelength division multiplexing (DWDM), such as optical filters, optical add-drop multiplexers (OADM), and others. However, temperature variation affects the optical and the mechanical properties of materials and thus the propagation characteristics of optical signals. Similarly, thermal expansion affects the properties of diffraction gratings, which is manifested as spatial angular shift. This shift has an adverse effect on the proper operation of grating devices and the function they perform. In this paper, we provide a temperature sensitivity analysis of diffraction gratings and a self-compensating method that provides performance stabilization due to temperature.

This communication primarily deals with utilizing organic electro-optic (OEO) materials for the fabrication of active wavelength division multiplexing (WDM) transmitter/receiver systems and reconfigurable optical add/drop multiplexers (ROADMs), including the fabrication of hybrid OEO/silicon photonic devices. Fabrication is carried out by a variety of techniques including soft and nanoimprint lithography. The production of conformal and flexible ring microresonator devices is also discussed. The fabrication of passive devices is also briefly reviewed. Critical to the realization of improved performance for devices fabricated from OEO materials has been the improvement of electro-optic activity to values of 300 pm/V (or greater) at telecommunication wavelengths. This improvement in materials has been realized exploiting a theoretically-inspired (quantum and statistical mechanics) paradigm for the design of chromophores with dramatically improved molecular first hyperpolarizability and that exhibit intermolecular electrostatic interactions that promote self-assembly, under the influence of an electric poling field, into noncentrosymmetric macroscopic lattices. New design paradigms have also been developed for improving the glass transition of these materials, which is critical for thermal and photochemical stability and for optimizing processing protocols such as nanoimprint lithography. Ring microresonator devices discussed in this communication were initially fabricated using chromophore guest/polymer host materials characterized by electro-optic coefficients on the order of 50 pm/V (at telecommunication wavelengths). Voltage-controlled optical tuning of the pass band of these ring microresonators was experimental determined to lie in the range 1-10 GHz/V or all-organic and for OEO/silicon photonic devices. With new materials, values approaching 50 GHz/V should be possible. Values as high as 300 GHz/V may ultimately be achievable.

We describe an in-service dispersion monitoring technique which we call dithered-filter method and show its application to adaptive dispersion compensation. In the proposed method, we filter out a portion of the received signal light by a dithered optical band-pass filter and derive the accumulated dispersion from the phase shift of the extracted clock signal. This method does not require any modification of the transmitter, and is applicable to any modulation format as far as the clock signal is extracted. In addition, owing to the dithering and the synchronous detection, high accuracy is realized with a short measurement time less than 200 msec. In this paper, we explain the principle of operation of the proposed method and show the experimental demonstration of fast adaptive dispersion compensation in a 40 Gbit/s reconfigurable optical network.

In this work, an optical short pulse generator is designed consisting of a pulse compressor and cascaded notch filter type repetition rate doublers. The performance characteristics such as pulse width and peak power as a function of design parameters are studied. The pulse compressor is optimized based on the simulation results. The 6ps wide pulses at 20 GHz repetition rate directly generated from mode-locked fiber laser (MLL) are compressed to 1.25ps wide pulses. Using a set of polarization maintaining fiber (PMF) loop mirrors the repetition rate is quadrupled and stable 1.45 ps wide pulse train at 80GHz is achieved.

In this paper, we describe the design theory for the supercontinuum spectrum generation in an optical fiber. To generate a wideband supercontinuum spectrum, the balance between fiber nonlinearity and the amount of group velocity dispersion is important. Secondly, the experimental results of supercontinuum generation are shown. A few kinds of optical fibers such as a highly nonlinear dispersion-shifted fiber and a highly nonlinear bismuth-oxide fiber are tested. Finally several applications of supercontinuum light are described. We demonstrate multi-wavelength light source, wavelength conversion, multiplexing format conversion, and optical characterization using a supercontinuum light source.

In this work, we demonstrate clock recovery from a patterned 160Gb/s optical-time-division-multiplexed (OTDM) return-to-zero (RZ) data stream. A cascaded LiNbO₃ Mach-Zehnder modulator is employed as an efficient optical-electrical mixer. A phase-locked-loop (PLL) is used to lock the cross-correlation component between the optical signal and a local oscillating signal. As a result, clock signal at 10GHz is extracted from the 160Gb/s optical TDM signal. The measured root-mean-square (RMS) timing jitter of the 10GHz clock signal is ~ 130fs.

The effective demultiplexing of WDM signals requires a tuneable filter, which is able to arbitrarily select and filter more than one channel to a particular output at any one time. Thus, such optical filters need to posses a large dynamic tuning range, narrow bandwidth and high sidelobe-suppression capabilities. Important applications of these tuneable filters include Add/Drop multiplexers and wavelength-selectors for tuneable lasers. The proposed architecture consists of cascading a Polarisation Independent Acousto-Optic Filter (PIAOTF) with a Fabry Perot Interferometer (FPI). The frequency domain characteristics of this architecture outperforms those currently offered by tuneable technologies, including the Fibre-Bragg Grating, the Mach-Zehnder Interferometer, the Fabry-Perot Filter and the Double-Stage PIAOTF. Consequently, our architecture is very favourable for WDM purposes since the filter boasts a sidelobe-level attenuation of <-25dB when simultaneously switching four channels, whereas the leakage at the through-port is only -16dB. In addition, the filter proves to demonstrate a convincingly flat passband with a 3dB bandwidth of ~0.35 nm. Consequently, a simple model is built to simulate the WDM (demultiplexing) environment in which the novel tuneable filter may be deployed. In this case, the tuneable filter's architecture enables a light stream consisting of multiple wavelength-channels, each with a spectral width of 0.1nm, to be split into two output ports via its bandpass and notch spectra. Hence, there is both a filtered and non-filtered light stream.

We report on an Optical Add/Drop Multiplexer designed from a single Arrayed Waveguide Grating and a MEMS optical switch. Three different MEMS switches were tested, the most promising being a potentially integrable binary actuator array.

All optical Boolean logic functionality has been demonstrated experimentally using integrated semiconductor optical amplifier (SOA) based interferometry at 40 Gb/s. The demonstrated functions are XOR, AND, and OR. The performance of the operations has been analyzed by solving the rate equation of the SOA numerically. The high-speed operation is limited by the carrier lifetime in the SOA. In order to solve the limitations imposed by carrier lifetime, a differential scheme for the XOR operation has been experimentally investigated.

A numerical model of wavelength converters based on cross-gain modulation in semiconductor optical amplifiers is developed. A more accurate gain coefficient model, spatial variations of carrier densities, and position-dependent differential gain and internal loss are included.

We report the design and fabrication of a high speed surface illuminated pin photodetector with a wide spectral response. An InGaAs based detector was grown lattice matched to InP and the device was fabricated using a novel technique to facilitate the direct absorption of incoming photons in the InGaAs layer without being absorbed by any other wider bandgap material. The absorption of a wide spectrum of wavelengths was achieved by recess etching almost all of the InP contact layer above the InGaAs absorption layer of a pin photodiode subsequent to ohmic contacts formation. Theoretical simulation shows responsivities above 0.6 A/W between 900 and 1600 nm and a linear reduction to 0.3 A/W at 650 nm. This makes the detector operational in both visible and near-infrared spectrum. The responsivities are 0.50, 0.77, and 0.67 A/W for 840, 1310 and 1550 nm respectively. These values confirm the potential of the device to be effective in all of the optical fiber communication wavelengths. Our calculations also show that the device can operate above 10 GHz throughout the spectrum.

Very low propagation losses in straight planar photonic crystal waveguides have previously been reported. A next natural step is to add functionality to the photonic crystal waveguides and create ultra compact optical components. We have designed and fabricated such structures in a silicon-on-insulator material. The photonic crystal is defined by holes with diameter 250 nm arranged in a triangular lattice having lattice constant 400 nm. Leaving out single rows of holes creates the planar photonic crystal waveguides. Different types of couplers and splitters, as well as 60, 90 and 120 degree bends have been realized. We have designed and fabricated components displaying more than 200 nm of useful bandwidth around 1550 nm. Design strategies to enhance the performance include systematic variation of design parameters using finite-difference time-domain simulations and inverse design methods such as topology optimization. We have also investigated a new device concept for coarse wavelength division de-multiplexing based on planar photonic crystal waveguides. The filtering of the wavelength channels has been realized by shifting the cut-off frequency of the fundamental photonic band gap mode in consecutive sections of the waveguide. Preliminary investigations show that this concept allows coarse de-multiplexing to take place, but that optimization is required in order to reduce cross talk between adjacent channels and to increase the overall transmission. In this work the design, fabrication and performance of these planar photonic crystal waveguide components are reviewed and discussed.

Photonic crystal waveguides have long attracted much attention in the integrated photonics community due to their high confinement properties and potential for the achievement of photonic circuits with a very high level of integration. While high propagation losses still impair most of the practical applications of such waveguides, predicted and demonstrated slow and dispersive propagation within compact lengths remain very attractive for optical signal processing. In this talk, results will be presented from an investigation on slow and dispersive propagation in two different types of InP-based photonic crystal waveguides fabricated at UCSB. Waveguides of the membrane type, with very strong vertical confinement, were fabricated and characterized, as well as guides with weak vertical confinement and deeply-etched holes. Those of the latter kind were successfully integrated with structures found in standard photonic circuits produced in our group. Detailed measurements of transmission will be presented showing slow and dispersive propagation close to band edges. Reasonable group delay enhancement is found, which is clearly dependent on propagation losses; on the other hand, extremely large GVD is found over reasonably wide bandwidths, even when considerable losses are present. This suggests that, by proper tuning of coupling coefficients, very compact dispersion-compensating elements can be designed. A discussion on the advantages and disadvantages, as well as different possibilities of using this class of waveguides for the implementation of delay lines and dispersion compensation will be presented.

We present the current status and developments of a horizontal beam path laser propagation experiment over the sea performed off the coast of Puerto Rico. Atmospheric turbulence effects have been measured by a Shack-Hartmann wavefront sensor with a Dalsa CCD camera and by a scintillometer from Optical Scientific, Inc* (OSI). We present preliminary scintillation measurements for an approximate period of two days from the two optical systems during the month of July 2005, also suggestions for improvement in the software, data acquisition protocol and hardware are presented.

We give a brief overview of some basic concepts in quantum information processing, and discuss various systems considered to implement these ideas. We focus our attention on one particular system, ultracold Rydberg atoms, and show how the strong interaction between such atoms can be used to implement an universal quantum logic gate, namely a phase gate. We also describe how quantum information can be stored and manipulated in mesoscopic ensembles of such atoms. We present some experimental results on the effects of strong interactions between Rydberg atoms.

Quantum cryptography applies the uncertainty principle and the no-cloning theorem of quantum mechanics to provide ultra-secure encryption key distribution between two parties. Present quantum cryptography technologies provide encryption key distribution between two parties. However, practical implementations encryption key distribution schemes require establishing secure quantum communications amongst multiple users. In this talk, we survey some of the state of the art quantum encryption deployment in communication networks. We will also discuss some common topologies that are being considered for multi-user quantum encryption networks. The performance of the multi-user quantum key distribution systems is then compared for four different optical network topologies: the Sagnac-based fiber ring, the wavelength routed, the passive star and the bus network. Their performances are compared and analyzed using quantum bit error rate analysis.

Optical spectroscopic properties of Er³⁺ and Tm³⁺ ions in new niobium tellurite glasses, Te0₂-Ba0-Sr0-Nb₂0₅ (TBSN), were investigated for (S+C)-band optical amplifications. On the basis of consideration for cross-relaxation among Tm³⁺ions and multi-phonon relaxation, it was shown that the presence of Nb₂0₅, which was source of the highest energy phonon, contributed to relaxation processes in TBSN glasses. The C-band amplification transition for Er³⁺ and the S-band amplification transition for Tm³⁺ are both sufficiently efficient in TBSN glasses and the Er³⁺-Tm³⁺ co-doped TBSN fibers could be efficient seamless (S+C)-band amplification media by selecting the concentrations of Er³⁺ and Tm³⁺

In this paper, we investigate a SOA (semiconductor optical amplifier) preamplifier model by suitable choosing the material and device parameters for the SOA to reduce the amplified spontaneous emission (ASE) noise at low range of 22.3 μW with 0.1 mW input power for PIN receiver. Our proposed SOA optical preamplifier is found to be more relaxed from optical alignment & antireflection coating and eliminate the need of optical filter and give large tolerance of an input light wavelength more than 100 nm. Also we investigate the receiver sensitivity for different bit rate. We show that receiver sensitivity is -69.9 dBm at bit error rate (BER) of 4.6×10^-10 for 10 Gb/s and for the 40 Gb/s bit rate the improved receiver sensitivity is -19.2 dBm with PIN receiver. We also observe the sensitivity of -40.5 dBm at 40 Gb/s with DPSK (differential phase shift keying) receiver. Further the impact of amplified spontaneous emission power, gain variation with input light wavelength & optical gain for TE & TM modes with input power for PIN & DPSK receivers has been illustrated. We show that optical gain of SOA is polarization independent and remain constant 30.06 dB up to gain saturation. We also show that DPSK receiver suffer less from ASE noise. The variation of material loss, length & bias current for our SOA preamplifier to optimize the receiver sensitivity is further illustrated.

In the optical interconnection of board level, 45-degree micro mirrors are used to achieve 90-degree optical path change. These mirrors are often fabricated by using a rotating blade and this method has a serious problem of cutting other waveguides in the vicinity. A self-written waveguide ha been successfully applied to repari and connect these waveguides. So, it will be possible to allow optical signals to be input and output at a specific location on the board.

We have developed "tapered self-written waveguide (SWW)" for coupling of optical components in the board level optical interconnection. Tapered waveguides have a possibility of achieving larger alignment tolerance for optical coupling. The fabrication condition of tapered SWW was studied by adjusting both the optical power and irradiation time of curing resin, and tapered SWW was successfully realized. The optical tolerance vertical to the optical axis twice as compared with straight SWW was obtained.

Photonic devices based on novel bromo-fluorinated poly(arylene ether ketone)s have been prepared. In these materials, the intrinsic optical losses at 1550 nm, due to the absorption of hydrocarbon bond overtone vibration, have been minimized by replacing H in the C-H bonds with Br or F, thereby shifting the overtone absorption to a longer wavelength. Typical optical slab losses of these materials are ~ 0.5 dB/cm at 1550 nm. In addition these materials have high thermal stability (5 wt% loss at temperature greater than 450°C), and are easily processed at temperatures lower than those previously reported for other poly(arylene ether)s (< 200°C). High quality waveguides have been fabricated using standard photolithographic processes. A thin film of silicon dioxide deposited by rf sputtering or e-beam evaporation on the polymer surface was used as a mask for reactive ion etching. Data on the design, fabrication and characterization of wide-band wavelength division multiplexers are reported. The devices exhibit on-chip losses of 7 dB including the fiber to chip coupling loss, output uniformity of ± 0.5dB and central wavelength thermal sensitivity lower than 0.06 nm/^oC. Optimization of devices through material properties and fabrication process parameters is discussed.

A finite element method with artificial transmitting boundary incorporated is firstly applied for acousto-optic tunable filters (AOTF) with weighted coupling. Here, two kinds of weighting function are considered, one is Cosine function, the other is Gaussian function. And the numerical results are shown for AOTF with a tapered acoustical directional coupler with different weighing function on a LiNbO₃ substrate. Through comparison, the optical filter responses of the TE-TM mode conversions of AOTF with Gaussian weighting function are better than AOTF with Cosine weighting function. The results are in good agree with others, which illustrates this method is useful in calculation on AOTF.