Making the middle arm of parallel arranged 3x3 fiber coupler feedback, we can gain a kind of fiber coupler which couple ratio can be controlled. If a parallel arranged 3x3 fiber coupler is not completely parallel, as long as the coupling coefficient k₁ between two adjacent fibers is greater than two times of the coupling coefficient k₂ between two not adjacent fibers F₁ and F₃, couple ratio can be controlled efficiently.

The operating characteristic of the semiconductor optical amplifier based Mach-Zehnder interferometer (SOA-MZI) all- optical add/drop multiplexer (AO-ADM) for optical time division multiplexing (OTDM) networks is analyzed in this paper. Based on the analysis of the crosstalk, the (pi) and -(pi) demultiplexing point operation are analyzed and compared. We find that, for the traditional symmetrical operation mode of the SOA-MZI, -(pi) demultiplexing point operation is favorable to have relative good extinction ratios for two output ports simultaneously. However the extinction ratio of the dropped and transmitted channels are still very limited arising from the gain dynamics of SOA and the symmetrical configuration of MZI employed. Besides, studies show that there is a large amplitude modulation (AM) in the transmitted channel originating from the gain dynamics of the SOA. To tackle these problems, the asymmetrical operation mode is proposed. Analysis shows the asymmetrical operation mode can improve the extinction ratios of the dropped and transmitted channels by optimizing the splitting ratio and the relative small signal gain ratio of the two SOAs. The AM of the transmitted channel for the asymmetrical operation is also less than that of the transmitted channel for the symmetrical operation mode.

The next generation fiber optical communication systems can be characterized in three trends: higher speed, better spectrum efficiency and longer transmission reach. To meet the requirement of next generation systems, passive DWDM components must have the following characteristics, low dispersion, small channel spacing, low insertion loss and high isolation. In this paper, we discuss one of the most important passive components: interleaver. Interleavers have been widely used in fiber optical communications systems. Their main functions include mux, demux, optical add/drop and bi-directional noise reduction. As data rate of fiber optical communication systems increases to 10Gb/s or higher, chromatic dispersion (CD) of interleavers becomes increasingly important. It has been shown that chromatic dispersion of interleavers could significantly limit the transmission reach. In this paper, we discuss 25GHz and 50GHz interleavers and their applications. Chromatic dispersion of several interleavers available on the market are discussed. We explain the generic relation between CD and channel spacing. The CD and CD slope of some interleavers can be reduced or completely compensated by innovative designs. We present an interleaver design with extremely low CD and CD slope. Finally, we present a 10Gb/s 160-channel transport application by using the low-CD 25GHz and 50GHz interleavers. The eye diagram and bit error rate of the 10Gb/s system will be discussed.

We show that a dual-core fiber can function as a wavelength selector for applications in Dense WDM systems. The fiber consists of an Erbium-doped core and an undoped core. The principle of operation relies on the large nonlinear effect of the Er-doped core when it is pumped by a 980nm laser diode. The selection of wavelength is achieved by adjusting the pump power launched into the doped core. Wavelengths with spacing of 1nm can be selected with crosstalks as low as 35dB.

A novel Newton recursive algorithm is proposed for an optimum design of arrayed waveguide gratings, which is different from the traditional complicated power-series expansion of the light-path function. The structure of an arrayed waveguide grating is represented by three constraint equations which may be chosen to meet some specific design demands. The new algorithm combines Newton algorithm with structure nonlinear constraint functions, which makes it more general and flexible for the optimum design of the device. From the initial value given, the arrayed-waveguide positions and matched waveguide lengths are determined from the numerical solutions for the roots of three constraint equations through a Newton recursive procedure in sequence. Anastigmatic mounts of arrayed waveguide gratings based on this algorithm are processed, and a three stigmatic-points one is designed. Further applications of this algorithm are also discussed, including the one that can not be designed with the theory of the power-series expansion of light-path function.

Fast-growing internet traffic volume require high data communication bandwidth over longer distances. Access network bottlenecks put pressure on short-range (SR) telecommunication systems. To effectively address these datacom and telecom market needs, low-cost, high-speed laser modules at 1310 to 1550 nm wavelengths and avalanche photodetectors are required. The great success of GaAs 850nm VCSEls for Gb/s Ethernet has motivated efforts to extend VCSEL technology to longer wavelengths in the 1310 and 1550 nm regimes. However, the technological challenges associated with materials for long wavelength VCSELs are tremendous. Even with recent advances in this area, it is believed that significant additional development is necessary before long wavelength VCSELs that meet commercial specifications will be widely available. In addition, the more stringent OC192 and OC768 specifications for single-mode fiber (SMF) datacom may require more than just a long wavelength laser diode, VCSEL or not, to address numerous cost and performance issues. We believe that photonic integrated circuits (PICs), which compactly integrate surface-emitting lasers with additional active and passive optical components with extended functionality, will provide the best solutions to today's problems. Photonic integrated circuits have been investigated for more than a decade. However, they have produced limited commercial impact to date primarily because the highly complicated fabrication processes produce significant yield and device performance issues. In this presentation, we will discuss a new technology platform of InP-based PICs compatible with surface-emitting laser technology, as well as a high data rate externally modulated laser module. Avalanche photodetectors (APDs) are the key component in the receiver to achieve high data rate over long transmission distance because of their high sensitivity and large gain- bandwidth product. We have used wafer fusion technology to achieve InGaAs/Si APDs with much greater potential than the traditional InGaAs/InP APDs. Preliminary results on their performance will be presented.

Nonlinear fiber optics, in the form of stimulated Brillouin scattering (SBS), has now emerged as the essential means for the construction of active optical devices used for all-optic in-line switching, channel selection, amplification, oscillation in optical communications, optical logic elements in optical computation and sensing, and a host of other applications. This paper attempts to present a survey and some of our own research findings on the nature of stimulated Brillouin scattering in single mode optical fibers and its device applications. In theory, the backscattering nature of the phenomenon enables its application as channel selectors and switches and filters in optical transmission and communications. We have been engaged in the design and implementation of fiber configurations, such as rings and loop mirrors, with the purpose of lowering the threshold. We report on experimental schemes involving Brillouin ring with amplifier-in-the ring, and Brillouin-NALMs (nonlinear amplifying loop mirrors). These successful devices are being studied for application as optical logic and neuron elements, Brillouin-NALMs for optical switching, and highly versatile sensors.

The Fabry-Perot laser diode coupled with external erbium doped fiber (EDF) terminated with a reflector forms the hybrid diode/EDF laser. The Fabry-Perot laser plays a role of spectral etalon with internal controlled gain for the linear erbium doped fiber laser when the main spectrum of longitudinal modes of Fabry-Perot laser overlaps the main gainband of erbium doped fiber amplifier. This idea has been proved experimentally. The hybrid diode/EDF laser can oscillate in multimode regime and it can simulate multichannel coherent source for Wavelength Division Multiplexing when only free spectral range of the Fabry- Perot diode lasers overlaps with ITU channel spacing.

A novel linearization process in electroabsorption modulated laser (EML) is proposed and experimentally demonstrated. Dual-parallel modulation scheme is used to compensate the nonlinear component of the EML by controlling the DC bias voltages of the each EML separately. The simulations on the nonlinearity of EML and linearization process are performed in both time and frequency domains. The validity of the proposed modulation scheme is confirmed through the experiments. From a dual-parallel modulation experiment at 8GHz, a reduction of 23dB in IMD3 and the following increase of linear dynamic rage of 19.6dB are achieved compared to those of a single EML operation.

An improved coupled mode approach is presented for the analysis of plane-wave interaction with a spatially periodic permittivity dielectric layer. The method consists of two parts, in the first part the eigenvalues and eigenvectors of coupled mode system of equations are corrected to reduce the inherent errors, which arise by neglecting second-order derivatives. In the second part, a scattering matrix representation of cascaded systems is used to model the effect of boundaries. The problem under consideration is a 2-D planar grating structure illuminated by a TE polarized incident wave, but the method can be applied to other grating structures such as couplers, or any other structure for which the coupled mode approximation is used, similarly. Comparisons with results obtained using alternative methods are given to verify the accuracy of proposed method.

Interleaver allows existing DWDM filters designed for operation at wide channel spacing to be extended to DWDM system with narrow channel spacing. Birefringent crystal interleaver is based on birefringence of crystal and interference of polarized light. The principle of approaching the flattop passband spectrum shape is analyzed by Jones matrix. All fiber unbalanced Mach-Zehnder interferometer interleaver and Michelson interferometer with GT resonator interleaver are also presented. Passband center wavelength precision and chromatic dispersion are discussed.

The reflectors of the surface acoustic wave (SAW) filters affect the filter performance, such as the insertion loss. The Cascade-Matrix-Transform Method provides a solution for the trivial calculation of the reflection of the surface acoustic wave (SAW) filters. It is easy to combine the multiple-reflection effect and propagation attenuation by arranging the cell matrix. The results show the effects of number of reflectors and the optimal ratio of the grating period to wavelength. The appropriate grating period is the half wavelength of the operating frequency with an optimal value of number of reflectors. The whole reflection of grating filters with different reflection coefficients at different frequencies in conjunction with the physical explanation of multiple reflection effects are disclosed in this paper. This methodology can also be extended to other further grated type filters, such as microwave filters or optical filters.

We propose a novel monitoring method of gain tilt for DWDM (Dense Wavelength Division Multiplexing) submarine cable systems with simple configuration. This method can be applied to DWDM systems on in-service condition without any special devices and dedicated wavelengths. The relation between the measurement sensitivity and pre-emphasis is studied, and we can distinguish more than 4dB gain tilt variation in 25nm transmission bandwidth.

In this paper, atomic functional models for optical transport network are introduced. Based on the proposed atomic components for optical components, the paper addresses the method modeling optical network elements. The models of some key network elements (i.e. OXC) are also shown in the paper, and simulation results by using the created model are present also.

Dispersion-managed solitons and their applications in ultra- long-haul transmissions are presented. The advantages of dispersion managed soliton transmissions over those of ordinary soliton and chirped RZ formats will be discussed. New designs in dispersion mapping, amplification (both Raman and EDFA), and transmitters and receivers will also be presented.

In this paper we investigate the maximum unrepeatered and uncompensated transmission on standard single mode fiber (SSMF) with single-sideband (SSB) modulation. To increase the optical signal to noise ratio (OSNR) we use a backward pumping Raman source. With the combination of both the bandwidth reduced modulation scheme and the OSNR improvement with Raman amplification we can enlarge the error-free uncompensated and unrepeatered transmission length for SSB modulation up to 200km.

Supercontinuum (SC) is a very promising technology which is widely used in a great deal of optoelectronics fields. In this letter, a SC producing process has been numerically simulated . The effects of chirp of the pump pulse on the width of SC are analyzed . and a series of conclusions are mentioned to explicate how optical pulse chirp affects SC when the SC fiber is DSF. Especially , the chirp effects using dispersion decreasing fiber (DDF) as SC fiber and the evolutions of SC along DDF of different chirp pulse are analyzed and compared . The threshold chirp for SC producing is defined. The simulation results are theoretical analyzed and some useful conclusions are given.

We study the optical pulse dynamics in a transmission system with periodic variation of dispersion, using the variational approach (VA). The dependence of soliton parameters on dispersion map strength is examined. We find that there is a critical map strength above which finite energy solitons can propagate at zero and normal average dispersion. The existence of two branches of soliton solutions in the normal dispersion regime for different levels of the pulse energy is observed.

To meet the ever-increasing traffic demands and to reduce the equipment as well as operation costs, recent developments in fiber-optic communication systems are focused on four areas: longer reach, wider bandwidth, higher bit rate and the intelligence of the system. In this presentation, recent advances in optical fiber amplifiers will be reviewed with the emphasis on the attributes supporting system requirements in these four areas. Topics to be addressed include ultra wideband amplifiers and amplifiers with new bands, a system perspective of Raman amplification, management of gain ripple, and intelligent amplifier control.

A polymer optical fiber amplifier (POFA) with double cladding was successfully prepared, The input signal at 595-nm wavelength was amplified (gain 5.6 dB) when the pump power at 532-nm wavelength was coupled into the inner cladding of the POFA with 43-cm length.

In a L-band EDFA, not only the pump power conversion efficiency (PCE) can be improved but also the gain can be clamped by C-band laser injection. On one hand, the injected C-band laser compresses the possible leakage of amplified spontaneous emission (ASE); while on the other hand, the maximum inversion level is limited by the high emission coefficient when the amplified C-Band laser light becomes the dominant pump source, which means that the maximum signal gain is limited so that the gain can be clamped. The numerical and experimental results show that the increment of PCE requires only a certain injected light power, but the clamped gain and the dynamic input power range is dependent on both of the injected wavelength and power.

All-optical clock recovery is the key technology in 3R all-optical regeneration. In this paper, to reduce the pattern effect in a bit clock recovering using injectied mode-locked laser based on semiconductor optical amplifier (SOA), we propose a novel technology named nonlinear polarization rotation, which is proved to be very effective by numerical calculation.

A meta-heuristic method to design Erbium-Doped Fiber Amplifier (EDFA) Module for Dense Wavelength Division Multiplexing (DWDM) Network is proposed. By combining Simulated annealing (SA) and coupled mode theory, we obtain a promising method for the design of Long Period Gratings (LPGs) based gain equalizer for EDFA module. In C-band from 1528.8nm to 1562.3nm, we get the EDFA gain fluctuation within 1.1dB.

A delay-time tunable, actively mode-locked erbium-doped fiber ring laser is demonstrated. The phase and the delay- time of optical pulse-train with repetition rate of about 500 MHz can be adjusted with maximum tuning range of up to 340 degree(s) (~1.9(pi) ) and nearly 1 period.

Adopting a new classical approach, the additive noise power generated in optical amplifiers is calculated in terms of power spectral density. The classical formalism used combines a coruscular approach to a phase-amplitude description of the optical field. The noise contributions of the input field fluctuations, including zero-point fluctuations, and of the electron momentum fluctuations at optical frequency linked to the amplifier itself, are clearly identified. The excess of noise associated to coupling or built-in losses is determinated. The well-known result of the Amplified Spontaneous Emission (ASE) is obtained for the laser amplifiers. This description is applied to linear phase-insensitive amplifiers and to inhomogeneous, nonlinear phase-insensitive Raman amplifier, pointing out the effects of gain compression and gain distribution. This new approach makes possible the treatment of the squeezed-state of light and the quadrature reduced noise amplifications.

Optical fiber Raman amplifiers are well suited for application in optical DWDM systems and networks. Instantaneous pump depletion due to different pump power requirement of bit 1's and bit 0's will induce extra optical noise and cross talk. This paper analyzes this kind ofnoise source and presents simulation results of the dependence of optical SNR drop on fiber length, bit rates and channel numbers.

A semi-analytical expression of double backward Rayleigh scattering (DRB) noise in a backward-pumped distributed fiber Raman amplifier (B-DFRA) is derived. A receiver is assumed to follow the B-DFRA, and various electrical noise terms in the receiver are compared with the DRB noise term which is shown dominating under elevated Raman on-off gain. An equivalent receiver sensitivity is introduced as a comprehensive measure of B-DFRA performance.

To design gain characteristic of dual-pump fiber Raman amplifier, an effect optimal model is investigated under different conditions, such as maximal gain and minimal gain ripper. Impact of pumps interaction, pump wavelength selection, signals interaction and signal saturation are discussed.

Analytical expressions of transient phenomena in discrete fiber Raman amplifiers (FRAs) are presented in this paper. It is found that the recovery time in gain-saturated backward-pumped FRAs equals to the round-trip time of optical wave in the FRA, and is typically in the order of 10-1 ms. The power excursion amplitude and recovery time will increase dramatically in cascaded FRAs, leading to severe system performance degradation. Characteristics of transience in cascaded and gain-clamped FRAs and some design guidelines are also discussed based on these analytical expressions.

Analytical expressions of equivalent noise figure and other noise characteristics of distributed fiber Raman amplifier (DFRA) are derived, for the first time to the knowledge of the authors, including difference in polarization state and attenuation coefficients between pump and signal light(s) as well as multiple-wavelength pumping. Accordingly, their impact on DFRA performance and optical signal-to-noise ratio improvement ((Delta) OSNR) of a cascaded DFRA+EDFA is investigated. With increasing DFRA on-off gain, (Delta) OSNR increases first linearly and then logarithmically, leading to saturation. Results of this work agree well with existing work. Based on these results, some guidelines on DFRA design is proposed.

The performances of SCDCF-based and RDF-based 40Gb/s OTDM and WDM systems are compared by means of numerical simulation. Theory and numerical results were presented and discussed with 40Gb/s signal of different duty cycles. Cross-phase modulation effects in 40Gb/s WDM systems are showed for the first time. We did find that in single channel transmission small duty cycle signal shows superior performance. The problem of RDF based system is mainly due to large nonlinearity of the RDF fiber we choose to calculate. The residual dispersion remains potential problem even at very excellent slope compensation in 40Gb/s systems. RDF based system offers more robust performance than SCDCF at the same value of compensation rate due to its relative short SMF fiber in every span.

An effective way to evaluate the performance of PMD compensation, which uses a proposed polarization-mode dispersion (PMD) emulator, has been studied recently. In this paper, we first present a theoretical analysis on a special model of PMD-emulator (PMDE), then explain the first and second-order PMD effects, which are very useful to value the impact of high-order PMD on pulse. Numerical and experimental results show that PMDE can emulate the statistics property and cover the range of high-order parameter of the real fiber successfully. The impact of second-order PMD on pulse for 40Gbit/s optical fiber communication is also presented as the simulation results.

In this paper, the characteristics of higher-order PMD up to third order and its impact on transmission systems are mainly studied. Based on the concept of second order principal states of polarization (PSPs) put forward in the paper, the covered ranges of higher-order PMD parameters with the increasing instantaneous differential group delay (DGD) value are investigated by simulation. It shows that the upper covered range of PSP rotation rate decreases rapidly with increasing instantaneous DGD until the DGD reaches the PMD value of the fiber, and then continues to slowly decrease. This may cause serious signal distortion due to fast PSP rotation rate even DGD at the low value. Furthermore, to that of the second order PSP rotation rate and that of second order PMD vector magnitude, they increase simultaneously with increasing instantaneous DGD until the DGD reaches about 2 times PMD value of the fiber, then turn to decrease. This may has an impact on the performance of higher-order PMD compensators. At the same time, the simulation results show that higher-order PMD parameters can impose serious signal distortion on 40 Gbit/s systems and the higher PMD value, the higher higher-order PMD influence on signal distortion.

Orthogonal polarization multiplexing technologies has been applied to dense wavelength division multiplexing (DWDM) transmission systems to reduce the crosstalk between the adjacent channels. It is one of the important factors in the viewpoint of not only the transmission performance but also the designing terminal equipment to clarify the relation of the polarization state between adjacent channels. In this paper, polarization state dependency was investigated experimentally in the DWDM transmission performance by using 11.4Gbit/s with 0.3nm channel spacing. Furthermore, as the long-term transmission performance is an important factor in the point of the stability, the long-term performance has been measured in the orthogonal polarization state. We have confirmed that the orthogonal polarization multiplexing is useful for the transmission performance and have shown its stability by the long-term measurement through the experiments.

Dispersion Compensation has to be performed when 32x10Gbs long haul signals transmission in G.652 fiber. Optimum system transmission performance depends on the amount and distribution of dispersion compensation, laser performance and the nonlinear effects of fiber transmission, etc. System performance calculation under dispersion compensation is given, and system performance influenced by various factors under complete and residual compensation conditions is analyzed in this paper, with 6x22dB as an example.

Statistical PMD emulator was proposed and implemented by using two polarization beam splitters and an optical delay line controlled by a microprocessor. The generated PMD had a perfect Maxwellian distribution and very low background auto-correlation (~0%).

The following paper analyzes a dispersion management technique for improved transmission fiber performance using high-order mode optical physics technology. Test results indicate that high-order mode technology precisely manages chromatic dispersion to enable high-capacity, long-haul and ultra long-haul optical networks.

Based on the weak intensity/phase modulation transfer matrix H, a matrix V for intensity noise (IN) and phase noise (PN) evolution along nonlinear, dispersive and lossy fiber is derived, while matrix description makes it especially useful in noise evolution analysis in systems with chained optical amplifiers. Evolution of two noise sources, laser phase noise and ASE of EDFA, is examined in detail through V. IN, PN and optical field noise spectral densities are studied uniformly. Correspondingly, electrical signal-to-noise ratio (SNR_e), equivalent linewidth ((Delta) (upsilon) _eq and optical signal-to-noise ratio (SNR_o) are used to characterize system performance. It is found that for IM systems, an optimal P₀ corresponding to minimum SNR_e exists in positive dispersion fibers, and input optical power cannot be arbitrarily increased to improve SNR_e even though only noise is considered. On the contrary, in negative dispersion fibers, SNR_e is improved with increasing P₀ monotonously. For PM systems with negligible laser phase noise and regardless of the sign of fiber dispersion, an optimal optical power also exists to ensure minimum phase noise. SNR_o, a combined contribution from intensity and phase noise, is better in fibers with negative dispersion than that in fibers with positive dispersion under the same P₀.

A rigorous analytical transmission matrix H for optical wave with weak modulation, including intensity modulation (IM) and phase modulation (PM), is derived considering the fiber nonlinearity, dispersion and loss based on the physical process. Four elements of matrix H correspond to conversions of IM->IM, PM->IM, IM->PM and PM->PM, respectively and the nonlinear components are extensively investigated. Physical insight of transmission effects is obtained explicitly. Matrix description makes it especially useful in the design and analysis of chained optical amplifier systems.

In long-distance optical transmission systems at high-speed faster than 2.5Gbit/s, several transmission degradations need to be overcome, such as group velocity dispersion (GVD), polarization mode dispersion (PMD) in the fiber and the ASE accumulation noise employing EDFA, etc. The use of forward error correction (FEC) codes offers the potential to improve these degradations and to increase BER performance. In this paper, a new simple decoding algorithm for triple-error correcting BCH codes is proposed. Without complicated matrix-operation or division-operation or intricate iterative algorithm, the algorithm is high efficient and high-speed because of its simplicity in structure. Especially, the parallel structure increases the speed of coding greatly. The result of hardware simulation confirms that the algorithm is feasible completely. The algorithm can be used in the high-speed optical communication system and other high-speed communication system field.

We have analytically and numerically evaluated the XPM-induced crosstalk in the conventional single-mode fiber (SMF) network upgraded by using dispersion-compensating fiber (DCF). We assumed that DCF's are used as interstage components in dual-stage optical amplifiers. The result shows that, although XPM-induced intensity interference can be minimized by various dispersion deployment schemes, there exists an interference floor caused by XPM-seeded modulation instability (MI). Symmetrical dispersion compensation is the most promising scheme for upgrading the existing optical network, if the reduction of single-channel transmission penalty is considered together.

In this paper, we describe the applications of novel long period fibre gratings written by high-frequency CO₂ laser pulses to Er-doped optical fiber amplifiers (EDFAs). Such a type of gratings has been used for both ASE (Amplified Spontaneous Emission) noise reduction and gain equalization of EDFAs for the first time to our knowledge. The ASE noise of two EDFAs used as a pre-amplifier and an in-line amplifier has been successfully reduced from 4.0dB to ~3.5dB and from ~4.8dB to ~4.3dB with a small-signal gain increase of ~7dB for the in-line amplifier, respectively. In addition, the gain profiles of other two EDFAs have been flattened to +/- 0.5dB by using the LPFGs fabricated.

In this paper, we report novel long-period fiber gratings (LPFGs) fabricated by using an improved CO₂ exposure method, which is based on the thermal shock effect of high-frequency CO₂ laser pulses at several kHz. Compared with the method using low-frequency laser pulses at several Hz reported previously, this new technique is more effective due to the significant increase in the transient laser energy density focused on the fibre, which is likely to cause larger densification and residual stress relief and hence larger refractive index change in the fibre. The optical and mechanical characteristics of such a type of LPFGs have been studied by experiment.

A novel interleave filter using Mach-Zehnder-Based sampled fiber gratings was proposed and fabricated. The interleaver demonstrates channel spacing of 100GHz and isolation >15dB.

In this paper, the effects of walk-off in NOLM-based wavelength conversion is analyzed, as well as group velocity dispersion (GVD) and self-phase modulation (SPM). Fiber Bragg gratings (FBGs) are introduced into the NOLM, serving as narrow bandpass filters. Corresponding coupled mode nonlinear Schroedinger equations (CMNSEs) and equations of the filter transmission window are given. Meanwhile, the optimal parameters of the system. Finally, time-resolved numerical analysis of our scheme is performed. And it reveals that the introduction of FBGs has improved the performance of NOLM-based wavelength conversion.

Spurred by exploding demand for transmission capacity, several techniques have been developed in dense wavelength division multiplexing(DWDM) systems for packing as many channel as possible into fibre-amplifier active regions. An interleave filter, which separates DWDM signal channels with equal spacing into two groups with twice the spacing, has received much attention. To be effective, this filter must have a wide flat-top passband, low insertion loss, minimal crosstalk and low dispersion. A polarization-based interleave filter can be fabricated by using cascaded birefringent crystals. Filter of this type can provide superior peak flatness and isolation for narrow channel spacing over what can be obtained in traditional interferometric devices. In such filter, the crystals are cut with their optic axes perpendicular to their length. By using a method based on an analogy to the synthesizing of finite impulse response(FIR) digital filters, the angles of the optic axes and the length of the crystals can be determined. An example of the design of a three cascade birefringent crystals, 50GHz polarization-based interleave filter is described. The allowable parameter error for this filter is also discussed. Furthermore, a polarization-based filter, which can separate the upper and lower channels of an input DWDM signal, can be designed by using this synthesizing method.

The paper presents a new encode/decode scheme of OCDMA using optical scanning. After studying the principle of optical scanning encode/decoder, a new structure of optical scanning encode/decoder is proposed. It can be seen that all-optical spread spectrum OCDMA encode/decode can be realized by the proposed scheme with optical scanning. Consequently, it provides an approach of getting rid of the bottleneck of optical and electrical exchange. It also makes the integrating of OCDMA encode/decoder more easily.

This paper presents the design and development of a bit rate transparent 3R O/E/O transponder, which accomplishes full data rate transparency from 30Mb/s to 2.5Gb/s by signal processing in electronic domain. With the use of chipsets that perform clock recovery in several continuous bit rate ranges, we develop a clock and data regenerating circuit self-adaptive to the bit rate of input signal. Experimental results presented in the last section of this paper show us that a good bit rate-adaptive capability and good BER performance can be achieved.

We propose, based on our recent investigations of a novel type of optical birefringent filter (called intracavity birefringent filter), a wavelength selective configuration, which may be used for Dense WDM applications. We show that this structure, formed with a classical Lyot-Ohman filter inside a ring cavity, takes the advantages of two typical filters: Fabry-Perot filters and birefringent filters, and in consequence gives a quite good compromise among usual filters parameters such as the working range, the bandwidth and the crosstalk. The theoretical considerations have been summarized. The first experimental results by use of a tunable nematic liquid-crystal light modulator have been shown.

As multilayer film can greatly enhance magneto-optical (MO) effect, it has become a very promising structure used in Faraday rotator for optical isolator. The transmission characteristics, however, vary greatly in different order of the magneto-optical layers and dielectric layers. In this paper, different orders and structures, such as different material of gaps and the distance between the gaps and the stack of some structures, are discussed in order to have large rotation (45 degree), high transmittance and small size, all of which are very important in the made of Faraday rotator.

The specification of DWDM filter becomes higher with the development of WDM technology. The filter must have sharp slope, low insertion loss, qualified pass band width, stop band width and temperature stability. Suitable substrate and coating material should be selected firstly. And multi- cavity F-P filter was adopted to design the narrow band pass filter.

Dense Wavelength Division Multiplexed Optical networks using all-optical devices represent the promising solution for future high-capacity wide-area network applications. One of the crucial factors which will determine their reliability and functionality is the performance of optical components. In this paper, Dense WDM Modules and Add/Drop Optical Modules are mainly considered. This paper discusses not only the theoretical advancement, but also the actual disposal in manufacturing those key DWDM components. We make long length of discussion on the optimum parameters, the realizable methods as well as testing consideration. All ideas are supported by practical results of experiment.

In this paper, we investigate the acceptance bandwidths due to different polarization of fundamental waves for type I quasi-phase-matched (QPM) second-harmonic-generation (SHG) in bulk periodically poled lithium niobate (PPLN) theoretically, which are determined by the departures from ideal quasi-phase matching in periodicity, wavelength and temperature. This analysis is useful for establishing fabrication tolerances for practical QPM bulk PPLN device. The numerical comparisons of acceptance bandwidths between type I QPM SHG E_z^(omega )->E_z^2(omega ) and E_y^(omega )E_y^(omega )->E_z^2(omega )(oo->e) at 150 degree(s)C in bulk PPLN are first presented. The grating periods for two various polarization interactions are given. The result shows that the acceptance bandwidths can be significantly enhanced by the use of the perpendicular polarization. Furthermore, the limitation of input fundamental wavelength for E_y^(omega )E_y^(omega )->E_z^2(omega )(oo->e) QPM SHG in bulk PPLN is discussed.

Using a spectrum-sliced device composed of two cascaded fiber Bragg gratings, a novel multiwavelength fiber light source is manufactured based on broadband amplified spontaneous emission from Er-doped fiber. Two channels incoherent light near 1550nm with FWHM of 0.23nm, 0.45nm, respectively, output power of above 1mW are obtained. Using this fiber amplifier light source, we transmit a 34Mb/s incoherent light signal over 10 Km single-mode standard fiber. The error floor level is less than 10^-9.

In this paper, one method for accurate evaluating EDFA small signal gain spectrum is designed. According to this method, EDFA can be regarded as a box. One can measure and determine its gain spectrum accurately by using a single continuous wavelength DFB laser and a wide spectrum light source.

Using a comprehensive model that includes pump-to-pump, pump-to-signal, signal-to-signal interaction, Rayleigh scattering and amplified spontaneous emission the Raman gain spectra of multi-pumped Raman amplifier under the various conditions were investigated. The results show that a different pump powers and pump wavelength intervals schemes are required for multi-pumped Raman amplifier. To obtain broader bandwidth and smaller gain ripple the powers of the shorter and longest wavelength pumps are needed higher power and the powers of the middle wavelength pumps should be set lower.