This PDF file contains the front matter associated with SPIE Proceedings Volume 7136, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and the Conference Committee listing.

Short locking and unlocking time is essepecially important in the packet clock recovery, as it determines the network resource utilization. This paper presents a novel 10-GHz all-optical packet clock extractor with ultrafast locking and unlocking time via self-phase modulation (SPM) effect of the semiconductor optical amplifier (SOA). Low finesse Fabry-Perot (F-P) filter and a saturated SOA, in cascade, are used to directly extract the packet clock from the data stream. An optical bandpass filter(OBPF) is used to filter out the red-shifted parts of the packet clock in order to reduce the locking and unlocking time. We analytically investigate the impact of the locking and unlocking time thanks to the self-gain modulation (SGM) effect and self-phase modulation(SPM) effect induced by the SOA, and demonstrate 10GHz clock extraction with the establishing time of 200ps and vanishing time of 600ps experimentally.

We propose an advanced structure of optical coupling which realizes high coupling efficiency and simplifying alignment, enabling the application of optical interconnect. A high-efficiency optical transmitter module based on this advanced coupling was fabricated using a VCSEL array. The VCSEL was directly bonded on bracket which is pasted on a holder with two pin used to locate. And there is a dummy club at the middle of the package for laser driver, favorable for cooling and reducing the noise caused by ground return. Coupling component were fabricated. And bending loss was tested. A successful eye diagram at the speed of 3.3Gb/s/ch with 850 nm was accomplished from the VCSEL-directbonded transmitter module.

A novel tunable and reconfigurable microwave photonic filter based on two cascaded Mach-Zehnder modulators (MZMs) and a dispersive medium is presented, theoretically discussed and experimentally demonstrated. A single-wavelength laser diode (LD) and the first MZM are used to obtain a multi-wavelength optical source, and the second MZM is modulated with the signal to be processed. The dispersive medium provides the time delay for different taps, since different wavelength experiences different time delay, when it travels in the optical fiber. By adjusting the modulation frequency and the bias voltage on the first MZM, the wavelength spacing and the relative amplitude of the optical tones after the first MZM can be changed, thus making the filter tunable and reconfigurable. The presented microwave photonic filter has been implemented in the experiment. The experimental results agree well with the simulation results, and show that the present microwave photonic filter has good tunability and reconfigurability. The mainlobe-to-sidelobe ratio (MSR) of around 25dB is achieved for the presented filter in the experiment.

A scheme for mitigating patterning effects in wavelength conversion by using a concatenated semiconductor optical amplifier (SOA) and electroabsorption modulator (EAM) is proposed. The scheme is investigated theoretically and experimentally. The bit error ratio and the power penalty of the converted output signal are estimated and compared with the results of wavelength conversion based on cross gain modulation in a single SOA.

Dispersion-induced radio-frequency (RF) power degradation in the optical true time delay (OTTD) module implemented by a broadband light source (BBLS) and a dispersion element is studied in this paper, both theoretically and experimentally. Theoretical results indicate that compared with that using a tunable laser source (TLS), OTTD using a BBLS suffers from extra power degradation, and that when the transfer function of the tunable optical filter (TOF) can be approximated to be rectangular, at additional frequencies besides the power-vanishing frequencies in the case of using TLS, the phenomenon of electrical power-vanishing can also be observed. Experiments using a TLS and a BBLS are performed for comparison, and TOFs of different bandwidth are used. The electrical transfer functions are measured and prove the conclusion of the theory. The electrical power-vanishing frequencies are provided and in good agreement with the theoretical values.

We show that a ring resonator with mutual modes coupling can achieve pulse delay or advancement of tens nanoseconds, which is similar as a ring resonator with single mode. Nevertheless, the pulse response can be sensitively tunable either through mutual mode coupling or through waveguide-ring coupling in the vicinity of resonant frequency.

Studied are the effects of fiber dispersion and loss and single-photon detector dark counts on key transmission rate and bit error rate of quantum key distribution system through theoretical analysis and computation. Research demonstrates that fiber loss is the main factor that influences quantum key transmission rate, while quantum bit error rate is related to fiber loss, dispersion, pulse width, dark counts and Rayleigh backscattering. Theoretical calculation indicates that in the current single-mode fiber-based plug & play quantum key distribution system with a dark counts of 10 ^-5/ns, the utilization of short pulse with 1550 nm wavelength and a 50 ps width can achieve the maximum transmission distance of 131 km. If we reduce dark counts or use G.653 dispersion-shifted fiber with short pulse, we can further improve key transmission rate and reduce the quantum bit error rate, thus effectively enhance the transmission distance of quantum key distribution system.

In this paper, the authors present and experimentally demonstrate an all-optical frequency up-conversion and demultiplexing method in a radio-over-fiber (RoF) system. A Mach-Zehnder modulator (MZM) and a microwave photonic filter based on a fiber ring structure are used to simultaneously realize the frequency up-conversion and demultiplexing functions. In this proposed system, only a fiber ring and a photodiode (PD) are needed in the base station (BS) to fulfill the signal processing functions. This simple solution makes the base stations more cost-effective, and shows good application potential in the future radio-over-fiber systems.

In this paper, the different properties of bursts of bypass traffic and bursts of local adding traffic were emphasized in Optical Burst Switching (OBS) networks. Based on this phenomenon, a hybrid analytic model of burst loss probability was developed. The simulation results show that our model is more accurate than the previous models.

A wavelength scheduling algorithm with limited-range wavelength conversion in OBS network is presented in this paper. Two wavelength conversion policies, First-Fit (FF) and Nearest Wavelength First (NWF) are adopted respectively in this scheduling algorithm. Numerical results show that the FF and NWF perform nearly identical in the scenarios of sparse wavelengths with small conversion range and large wavelengths with small conversion range. The FF outperforms NWF when the wavelength number is large and the conversion range is wide. However, consider the fact that output power strongly deteriorates as a function of the distance between the ingoing and outgoing wavelengths, the NWF policy should be adopted in the practical OBS networks.

For metro and regional 100-Gbps transmission, a transparent channel reach of 500-600 km is required and a 100-GHz channel grid is typically used. For these applications, a cost effective modulation format is introduced which can make use of electronic components designed for the already established 40-Gbps market, bypassing the requirements for novel electronic developments and therefore reducing the component cost. With this DPSK-3ASK modulation format, five information bits are transmitted in two consecutive symbols, leading to a symbol rate of 45 Gbaud, including overhead for framing and FEC. To minimize hardware requirements and to create a cost-effective solution, a single Mach-Zehnder modulator can be used to create the optical DPSK-3ASK signal after combining the phase and amplitude modulation signals into a 6-level modulator drive voltage. In this paper, it is demonstrated by numerical simulations that these voltage levels can be modified to adapt to varying signal distortions and thereby yield improved transmission performance. It is shown that by dynamically modifying the modulation levels based on the channel performance, dynamic signal impairments such as the non-linear effects from varying power levels, changes in chromatic dispersion, or varying PMD levels can be mitigated. Error-free performance (with FEC) can be obtained with 24 dB OSNR and 7ps DGD for a 112-Gbps (45-Gbaud) optical signal.

We introduce two important technologies for development of next generation ultra-high-speed optical communications: (i) polarization multiplexing, phase modulation with digital coherent detection, and (ii) OFDM-based optical fiber transmission. In both schemes, digital signal processing plays a key role in recovering the signal and mitigating the detrimental effects from optical signal transmission. We further describe a novel three dimensional low-density parity check (LDPC) coded modulation scheme, including its principle and system performance.

A novel approach for optical DPSK/PolSK orthogonal encoding based on four-wave mixing effect in a semiconductor optical amplifier is proposed and tested using VPI simulation platform. The combined DPSK/PolSK signal is successfully generated and demodulated with clear and open eyes.

This paper proposes and numerically investigated a novel high-speed wavelength-division-multiplexed passive optical network (WDM-PON) architecture with colorless user terminals based on the use of a different modulation scheme for downstream and upstream transmission. In the central office (CO), based on the carrier suppressing functionality of Mach-Zehnder modulator (MZM) and differential-phase-shift-keying to amplitude-shift-keying conversion using Mach-Zehnder delay interferometer, 40-Gb/s optical frequency shift keying (FSK) is generated and employed for transmitting the downstream data. In the remote node (RN) or optical network unit (ONU), the upstream data is re-modulated at 2.5-Gb/s by an intensity modulator and sent back to CO with the same fiber. Since only one light source is needed for each WDM channel, and the MZM to generate carrier suppressed signal can be shared by all the channels, a centralized 40-Gb/s WDM-PON access system with low-cost configuration is realized. Error free transmission over 20-km SMF can be observed for both downstream and upstream signals in our simulation.

Nonlinear Mach-Zehnder interferometers, when driven by a phase-locked local oscillator pump, provide phase-sensitive amplification to input signals. So phase-sensitive amplifier(PSA) can be used to realize phase regeneration of phase modulated signals. In the paper, we adopt kerr media-based nonlinear Mach-Zehnder interferometer to form a PSA to regenerate DPSK signals, and we propose a structure for the phase regenerating of DQPSK signals using two balanced PSAs. Theoretical analyses of all-optical phase regenerations of DPSK/DQPSK signals based on phase-sensitive amplification are presented. Simulations show that nearly ideal phase regeneration can be achieved for the DPSK/DQPSK signals.

We investigate the impact of the reflection in an RSOA-based WDM PON utilizing the spectrum-sliced ASE source as the seed light. Since the spectrum-sliced ASE source has sufficiently large bandwidth to suppress the optical beat interference (OBI) noise, we can achieve the high reflection tolerance in comparison with that in a conventional WDM PON using the remodulation scheme. In the experiment, we systematically investigate the dependence of the reflection tolerance on the operating conditions such as the optical power and bandwidth of the spectrum-sliced ASE source injected into the RSOA. In the case when the injection power into the RSOA is low, the performance of the upstream signal becomes vulnerable due to the reflection-induced incoherent crosstalk and parasitic laser oscillation. However, we can improve the reflection tolerance significantly by increasing the optical power and the bandwidth of the spectrumsliced ASE source. By optimizing these operating conditions, we can achieve an excellent refection tolerance greater than -15 dB.

86 Gbit/s hybrid polarization multiplexed OOK/DPSK signal is generated and transmitted at the same wavelength. Clear eye diagrams of demultiplexed OOK and DPSK signals are obtained after transmission over a 106.5 km SSMF fiber span.

Three different schemes to generate optical millimeter-wave and wavelength reuse for up-link connection in the radioover- fiber (ROF) systems have been theoretically and experimentally investigated. We have compared the performance of the three different schemes for the radio-over-fiber systems considering the cost and configuration of their architectures. A novel scheme to generate optical millimeter-wave and realize centralize lightwave operation in the radio-over-fiber (ROF) systems has been demonstrated. This scheme has shown high performance and low cost compared with the existing schemes.

We propose and demonstrate simultaneous multiple dense wavelength division multiplexing (DWDM) channels optoelectronic non-return-to-zero (NRZ) to return-to-zero (RZ) regenerative format conversions based on a single phase modulator (PM) and a fibre delay-interferometer (DI). The PM is driven by a local RF clock signal, and the DI with free spectral range (FSR) equals to the channel spacing is used to extract the chirps induced by the phase modulation, for all the channels at the same time. Since the original carriers are suppressed to some extent while the chirps are extracted, thus the NRZ-to-RZ conversions can be achieved with regeneration. The proposed multi-channels format conversions are successfully demonstrated at 16*10 Gb/s, with channel spacing of 100GHz. Bit error ratio (BER) measurements show 3.5 and 4.2 dB penalty improvements for 50 and 75 km transmission without dispersion compensation, respectively.

Abstract-We demonstrated 8 × 10 Gbit/s packets buffered in Dual Loop Optical Buffer (DLOB).Confirmed by experiments, the crosstalk and unbalanced power among different wavelengths which is the main restraining of the buffering time can be avoided by power equalization using Semiconductor Optical Amplifier (SOA). The buffered data packets can be retrieved after they are buffered for 5 cycles. The power penalty, the Extinction Ratio and the Signal to Noise Ratio are 4.2dB, 3.98dB, 5.47dB. The eye diagrams for 8 wavelengths signals are obtained.

We propose that double broadcast services can be overlaid over high-speed point-to-point downlink data in a WDMPON with source-free optical network units (ONUs). In the optical line terminal (OLT), a set of single-drive Mach- Zehnder modulators (MZMs) are driven by downlink point-to-point data to generate a differential phase-shift keying (DPSK) format. The downlink DPSK signals from different wavelengths are multiplexed and then fed to a following dual-parallel MZM (DPMZM) as a double broadcast services transmitter. The broadcast service_1 is an optical carrier suppression (OCS) format, while the broadcast service_2 is an inverse return-to-zero (IRZ) forma. After the transmission, at each ONU, the optical signals are separated by an optical filter. The filtered OCS signal is detected to retrieve the broadcast service_1. The DPSK/IRZ signals are split into three parts, one part is detected by an IRZ receiver to recover the broadcast service_2, the second part is detected by a DPSK receiver to retrieve the downlink data and the third part is re-modulated by the upstream amplitude shift keying (ASK). We also perform an experiment to verify the feasibility of the proposed scheme, where the power penalties of less than 1.5 dB are obtained after 25-km transmission with 1.25- Gb/s data rate.

Facilitated by variable differential-group-delay (DGD) modules, we propose a simple method to extend polarizationmode- dispersion (PMD) monitoring windows three times larger using either degree-of-polarization (DOP) for 40-Gbaud differential-phase-shifted-keying (DPSK) differential-quadrature-phase-shifted-keying (DQPSK) signals. Such approach can be further incorporated into PMD compensators.

We review our recent achievements in the optical performance monitoring techniques such as the optical phasor monitor and optical signal-to-noise ratio (OSNR) monitor developed for differential phase-shift-keyed (DPSK) and quaternary PSK (DQPSK) signals. The optical phasor monitor is a tool that can measure the in-phase and quadrature components of the optical field of the phase-modulated signal and display the measured results on the complex plane as a phasor. We implement such a phasor monitor by using an adjustment-free differential-phase demodulator composed of a 120-degree optical hybrid (which can be realized simply by using a conventional 3x3 optical coupler), and demonstrate its plug-and-play (i.e., phase-adjustment free and wavelength/polarization-independent) operation experimentally. We also show that the proposed phasor monitor is well suited for the use in the diagnosis of DPSK/DQPSK. The OSNR monitoring is another important function to evaluate the signal quality as well as the condition of the transmission link. We have recently proposed a new technique for monitoring the 'in-band' OSNR of DPSK/DQPSK signals. This technique estimates the OSNR by analyzing the radio frequency (RF) spectrum obtained by the self-heterodyne detection. In this paper, we briefly explain its operating principle and show the experimental results. Using this technique, we could accurately monitor the OSNR of the 10-Gb/s DPSK and 20-Gb/s DQPSK signals in a 640-km long transmission link. The results also show that the performance of the proposed technique is not sensitive to the effects of chromatic dispersion and polarization-mode dispersion.

Coherent detection is a new attractive detection method in optical transmission, especially in high capacity long-haul fiber communication systems. Most coherent detection systems use a local laser as the local oscillator rather than using an optical phase-locked loop (OPLL). In this case, the signal and local laser's wavelength must be as close as possible, and both should be stable. Wavelength stability of DFB laser is a critical factor for optical coherent detection and other applications. We use fiber couplers and a fiber phase shifter to buildup an optical 90o optical hybrid, by which the wavelength fluctuation is monitored with a very high sensitivity.

The main techniques for high bandwidth optical digital sampling are briefed. Potentialities and drawbacks of optical solutions are investigated and compared.

This paper presents the Direct Sequence Optical Code Division Multiple Access (DS-OCDMA) system using two configurations of the optical source. To encode/decode transmitted data, we used Superstructured Fiber Bragg Grating (S-FBG) as encoders/decoders. We consider three unipolar family codes which are the Prime Sequence (PS), Quadratic Congruence (QC) and the Extended Quadratic Codes (EQC) codes. In order to evaluate the performances of our system in term of Bit Error Rate (BER), we implement the Importance Sampling (IS) technique, which is a variant of the wellknown Monte-Carlo (MC) method. Our simulation results depict that EQC codes outperform QC and PS codes for the DS-OCDMA system using either coherent or incoherent source. We show also that using EQC codes with increasing the optical bandwidth and maintaining the electrical filter leads to improve the performance of incoherent system.

A hybrid 1D/2D en/decoding O-CDMA system is proposed to support multiple bit rates. To the best of our knowledge, a hybrid O-CDMA system with 2.5Gb/s (OC-16) and 155Mb/s (OC-3) transmission along the single mode fiber with length of 60km is first successfully demonstrated. In this experimental demonstration, phase coded SSFBGs with chip rate up to 320Gchip/s are applied as 1-dimensional en/decoders in 2.5Gb/s transmission while 2D en/decoders are applied in 155Mb/s real time video transmission. Error free transmission is achieved in 1D encoded 2.5Gb/s link; while 2D encoded 155Mb/s channel carrying video signal can be transmitted along the same fiber. This proposed hybrid 1D/2D en/decoding O-CDMA system can also be used for O-CDMA PONs to reduce the costs at the ONU side.

Different operation models of joint optical multi-band detection systems are provided here, corresponding to the different characters of detecting probabilities of subsystems. It leads to a higher over-all detecting probability. The joint optical detection system is able to be operated under different models when detection probabilities of subsystems are low or there is big difference among the probabilities. When all the probabilities of subsystems are low, the joint optical detection system directly conducts data fusing. If there is big difference among the probabilities of subsystems, information from the subsystems is chosen before entering the fusion center. In this way, the false-alarm probability of the joint optical detection system reduces obviously comparing with direct data fusion. So it also can reduce the calculation of system and improve the competence of system.

When increasing channel bit rate beyond 10Gb/s or when operating over fiber lines with sparse or no in-line dispersion compensation, Kerr-like non-linear effects can be considered as second order with respect to dispersive effects, because pulse broadening can expand over numerous neighbor pulses, before optical non-linear effects imprint their signature noticeably. To accurately emulate the interactions between pulses in this case, a few studies emphasized that Pseudo- Random Binary Sequences (PRBS) should be used, with exponential dependence of the required PRBS length on bit rate and accumulated dispersion. In this paper, we explain our strategy to numerically estimate the required number of random, noisy bits for Monte-Carlo simulations, and show that it weakly increases in presence of pulse to pulse correlations and commonly tolerated levels of non-linearities (i.e. leading to transmission penalties as high as 1.5dB, for reference BERs of 10^-2, 10^-3 or 10^-5) . Then we determine the actual required PRBS length that yields the same (sufficient) BER accuracy as the MC method. We demonstrate its actual dependence on BER, and show that MC theory provides a reliable upper bound in FEC-assisted, highly dispersive systems.

A practical single-scattering model of solar blind UV communication channel, which is suitable for NLOS digital communication is presented in the paper. NLOS transmission loss of UV channel is deduced by Lambert law . Calculation method of scattering phase function and integral of intersection volume between transmitting beam and receiving beam are described detailed in prolate spheriodal coordinate. Base on them, a practicable calculation method about signal to noise ratio (SNR) and bit error rate (BER) of UV communication systems is presented, which construct a bridge between system parameters and digital system performances in simulation of UV communication channel and systems .Under certain BER and typical modulations in UV communication , relationship between system parameters and system performances are discussed.

In this paper, the influence of the burstiness of IP packets on nonlinear effects in dynamical optical Networks such as XPM (cross phase modulation) and FWM (four-wave mixing) are investigated and simulated by adding revised Poisson distributed traffic parameters in the nonlinear Schrödinger equation (NLSE).From simulations, different eye diagrams with different IP traffic loads, different fiber input light powers are achieved. When the input power is bigger than 3 dBm with 40 channels or bigger than 5dBm with 16 channels, the effect of IP bursty traffic could depredates eye diagrams dramatically. The research results are useful for router finding and system design.

ww have presented and demonstrated the dual-wavelength signal storage in Dual-Loop Optical Buffer. The output equality of dual wavelength signals will decrease for unbalanced gain and phase shift. The problem can be resolved by power equalization using the saturation character of SOA. The data packets with2 × 2.5Gb/s can be buffered in DLOB for 16 cycles corresponding to 20 .The Extinction Ratio of the output packet is 8dB while the S/N is 8.1dB.

As a special case of continuous phase frequency-shift-keying (FSK), minimum-shift keying (MSK) exhibits some different properties compared with the traditional optical phase modulation formats, such as return-to-zero (RZ) differential-phase-shift-keying (DPSK) and differential-quadrature-phase-shift-keying (DQPSK). In this paper, we investigated the receiver performance degradation caused by laser frequency offset between laser frequency and delay interferometer (DI) phase, which is found to be the most critical impairment for the receiver performance in a practical optical phase modulated system. Results show that MSK system is about double times and six times more robust to frequency offsets than RZ-DPSK and RZ-DQPSK systems operating at the same 10Gb/s bit rate, respectively.

Compared with the optical communication system using intensity modulation with direct detection (IM-DD), the coherent optical system enhances the sensitivity and enables access to all the optical characteristics. In this paper, a novel scheme of coherent QAM transmission system based on heterodyne optical detection is experimentally demonstrated. In the proposed scheme, the intermediate frequency (IF) carrier modulation at double-sideband suppressed carrier (DSB-SC) mode at the transmitter side enables demodulation of QAM signal with heterodyne detection. Furthermore, the receiver is simplified by avoiding the use of high frequency broadband microwave devices due to the employment of IF carrier modulation. The performance of the proposed coherent QAM system is analyzed theoretically based on numerical simulation. The simulation results show that the improved structure of the proposed system can effectively depress the phase noise induced by laser. An experiment of 4-QAM transmission exploiting the proposed coherent heterodyne system is presented to justify the principle.

DWDM networks have been evolving to higher spectral efficiency. To make optical duobinary format, which is considered as a suitable modulation format in MAN DWDM networks, can work in 50GHz channel spacing 43Gb/s DWDM networks is very significant. In this paper, it is pointed out that the delay time in delay-and-add optical duobinary system is one of key factors which affect system performance. The suitable delay time for 43Gb/s systems compatible 50GHz with 100GHz channel spacing is obtained by system simulation. The simulation results show that 0.8 bit delay time is more suitable for 43Gb/s ODB DWDM system at compatible 50 with 100GHz channel spacing than 1 bit delay time because of its higher Q value and bigger dispersion and nonlinearity tolerance.

Within the realm of OFDM research, (peak-to-average power ratio)PAPR is an important object to be considered. Overlarge PAPR can deteriorate the linearity of the wireless systems. In this paper, we interpret the concept of PAPR in optical OFDM system and testify that the modulation method in sub-carriers and nonlinear effects can affect PAPR in the optical OFDM system. Through simulation, PAPR problem stand out while considering OFDM in optical transmission technology, so reducing it has become an important subject in optical OFDM system.

Optical technologies represent the main bet for future communication systems. Among the others, digital subsystems for optical processing are of great interest thanks to their intrinsic properties in terms of bandwidth, transparency, immunity to the electromagnetic interference, cost, power consumption, as well as robustness in hostile environment. Key basic functions are represented by logic gate, logic function, flip-flop memories, optical random access memories, etc.. Research in this field is in its very early stages even if some interesting techniques have been already theoretically addressed and experimentally demonstrated. In this paper we review the state of the art for all-optical flip-flop including different approaches such as fiber based, semiconductor amplifier based or waveguide and micro-ring solutions. Best result will be highlighted in terms of transition speed, switching energy, complexity and power consumption. Finally our best results are discussed.

We propose a novel scheme employing phase to amplitude modulation conversion and cross-gain compression in Semiconductor Optical Amplifiers (SOAs) to implement 2R regeneration at 40Gb/s. The resilience on the pumping power and filtering are analyzed.

An all-optical regenerator for Quadrature Phase-Shift Keying (QPSK) and Differential Quadrature Phase-Shift Keying (DQPSK) signals based on Phase-Sensitive Amplifiers (PSAs) is studied through numerical simulations. It is found that the regenerator can provide significant suppression of both amplitude and phase noise, and that the BER performance of the regenerated signals can be greatly improved.

We demonstrate dispersion monitoring of differential phase-shift keying (DPSK) modulated optical transmission system by measuring RF components of detected optical signal. The results show that this scheme can realize dispersion monitoring up to 4320ps/nm and sensitivity up to 15.5(ps/nm)/dBm. The effect of OSNR on the performance of the monitoring system are studied and discussed.

Progress towards the definition of next-generation passive optical networks (PONs) based on wavelength-division multiplexing (WDM) is reviewed and compared to emerging requirements. A key challenge is providing ultra-high (e.g. 10 Gbps) bandwidth for demanding users while cost-effectively supporting less-demanding users. A new approach is presented in which diverse bandwidth requirements are supported on a conventional WDM PON outside plant through the use of flexible wavelength sharing in the local office. An example is demonstrated experimentally showing that with 16 users per passive node, each wavelength can be shared by up to 16 users distributed across up to 16 PONs served by the same local office. Factors limiting sharing and throughput are discussed.

A hybrid optical access configuration for simultaneously providing wired and wireless services was proposed. The proposed system can also remote deliver doubled local-oscillator signal, which can be reused in the uplink. At the central station, the signals used for wireless links are subcarrier multiplexing and modulated as PSK format, and the wired baseband signals are directly modulated as NRZ format, then two signals are simultaneously generated by an external integrated modulator and transmitted on the same wavelength. The doubled local-oscillator signal is extracted in the wireless links, then used in the uplinks, the recieved eye diagrams are pretty well.

We demonstrate 300-km transmission of 10Gb/s data over SMF-28 fiber without DCF using a 4-channel tunable chirp managed laser and EDC chip at the receiver, which can fit into hot pluggable XFP modules.

This paper demonstrates bandwidth requirement for optical DPQSK transmitter using one Dual-drive Mach Zehnder Modulator (DDMZM) for 100Gb/s physical transmission. The result shows that at receiver bandwidth of 40GHz, NRZDQPSK signal requires 60GHz modulation bandwidth at least, while RZ-DQPSK scheme demands less bandwidth, about 40GHz for 50% duty cycle RZ-DQPSK signal.

In this paper, we highlight the potential of Coherent PDM-QPSK format for next-generation 100Gb/s transmission systems, through a record transmission experiment of 16.4Tb/s over 2550km, and in-depth experimental analyses of tolerance to joint PMD and non-linear effects, as well as robustness to typical constraints of terrestrial optical networks.

A 100-Gb/s high-speed optical transmitter is proposed and experimentally demonstrated. Based on frequencyquadrupling technique, two sub-channels with a fixed 50-GHz spacing are obtained from one laser source. Using returnto- zero differential quadrature phase-shift keying (RZ-DQPSK) modulation format and polarization multiplexing (PolMux), only low-speed electronic devices of 12.5 GHz are needed for the 100-Gb/s transmitter. This eliminates the need of ultrahigh-speed optoelectronic devices and thus greatly reduces the cost. The experimental results show that this transmitter can achieve good performance in dispersion tolerance of a 25-km single mode fiber (SMF). The performance of the generated signal is also verified through simulation with respect to chromatic dispersion (CD), polarization mode dispersion (PMD) and nonlinearity.

Transmission of eight channel 100Gbit/s CS-RZ DQPSK signals up to 1500km SSMF was demonstrated. This transmission system only used EDFA and did not adopt coherent detection and polarization multiplexing. The span spacing in this system varies from 80km to 100km to simulate the practical optical network. Keywords: Transmission system, modulation format

A novel scheme of one access point with multi-service wireless signals for radio over fiber systems is proposed. The generation of two 200Mb/s wireless signals with central frequencies at 17.5GHz and 20GHz carried on one optical wavelength is the preliminary result of the concept-proof experiment.

In this paper, a practical 40Gb/s 12:16 converter is implemented for VSR parallel optical transmission. The converter realizes the functions of mapping OC-768 frame to/from parallel optics. Using two chips of Altera FPGA and Agilent81250, a complete simplex communication experiment system is built. Detailed design such as frame synchronization, deskew algorithm and converter are presented. In addition, we design a SFI-5 interface verifying board to verify the deskew function. Testing results are also given illustrating that the converter works well.

A novel upstream data remodulation scheme with downstream ASK-DPSK modulation format is proposed for WDMPONs in this letter. And this kind of scheme is also experimentally demonstrated on one particular wavelength channel. Both the 2.5-Gb/s downstream optical DPSK signal and the 2.5-Gb/s downstream optical ASK signal are demodulated respectively at the receiving side, while part of the downstream carrier is directly remodulated by an optical intensity modulator for upstream transmission of 2.5-Gb/s data. System performance under different amplitude ratios of downstream optical ASK signal is investigated. It is shown that effective reception and robust performance can be achieved after having been transmitted for 25-km.

Differential phase shift keying (DPSK) data format is employed to combat noise in the bandwidth limited coherent time spreading (TS) optical code division multiplexing access (OCDMA) system. In addition, the performance of DPSK-OCDMA system with turbo coding is analysed and simulated. Theoretical results show that performance degradation due to bandwidth limitation could be effectively restrained by the DPSK-OCDMA scheme, and further performance improvement could be achieved by incorporating turbo coding into OCDMA system. It is confirmed that the number of active users is increased from K = 11 to K = 25 in bandwidth limited OCDMA system by employing DPSK format and turbo code for a given BER of 10^-6.

We theoretically and experimentally report a single-wavelength module of 1x4 wavelength selective switch (WSS) module based on the bi-axial MEMS mirror and thin film filter technology. With a cascaded structure of such WSS module, a novel 1xN WSS can be realized. The insertion loss of 3dB, extinction ratio > 50dB and hitless switching are achieved.

For the first time, we experimentally investigated and demonstrated waveband switching for 100 Gb/s DWDM system with high spectral efficiency of 3.4 b/s/Hz. 3 adjacent 25 GHz-spaced 114 Gb/s PolMux-RZ-8PSK channels are grouped as a waveband to be switched and transmitted through 4 WSS-based ROADM nodes. Error free transmission is achieved with little filtering effect observed even after 4 ROADM nodes. The transmission performance shows that waveband switching based on WSS can be used in high spectral efficiency DWDM systems. This can reduce the switching hardware cost and operation expense in highly dense WDM systems.

An all-optical switch is proposed and demonstrated using the nonlinear polarization rotation (NPR) with linear polarization maintenance arising in a semiconductor optical amplifier (SOA). By introducing optical control pulse into the SOA, the state of polarization (SOP) of the packet signal can be switched between the vertical linear SOP and the horizontal linear SOP on the Poicáre sphere (PS). Then a polarization beam splitter (PBS) is used to perform polarization-to-intensity conversion. Thus the input packet signal can be switched between the two output ports of the PBS. And an extinction ratio of switching operation larger than 20 dB is obtained.

The seamless integration of broadband optical and wireless access networks is considered to be a promising solution for next generation access networks which will provide high capacity and flexibility with lower cost. In such access networks millimeter wave (mm-wave) radio over fiber (RoF) system is a key enabling technology due to its its large bandwidth and short reach in atmosphere. In this paper, we will discuss some key technologies based on injection locked lasers for mm-wave RoF systems, including all optical generation of mm-wave signal, up- and down-conversion, single sideband modulation, and transmission.

We perform a study of extremely slow pulse propagation speed in densely-doped erbium fibers by using a detuned signal. A quantum coherence effect, coherent population oscillations, produces a very narrow spectral "hole" in the homogeneous broadened absorption profile of the erbium-doped fiber. The resulting rapid spectral variation of the refractive index leads to a large value of the group index. The interparticle interaction between the nearest neighboring active particles due to the high density of erbium ions is considered. The result shows that the nonlinear refractive index can be greatly increased which leads to the smaller group velocity and the larger time delay.

Broadband networks are among the fastest growing segment in telecom. The initial and still very significant push originated with xDSL technologies and indeed a significant amount of research and development is still occurring in this field with impressive results and allowing for a remarkable use of the installed copper infrastructure way beyond its originally planned bandwidth capabilities. However it is clear that ultimately a more suitable fiber based infrastructure will be needed in order to reduce both operational and network technology costs. Such cost reduction in inevitable as the added value to end users is only related to services and these cannot be priced outside a sensible window, whilst the related bandwidth increase is much more dramatic and its huge variability must be met with little or no cost impact by the network and its operation. Fiber in access has indeed the potential to cope with a huge bandwidth demand for many years to come as its inherent bandwidth capabilities are only just tapped by current service requirements. However the whole technology supply chain must follow in line. In particular optical technology must brace itself to cope with the required much larger deployment and greater cost effectiveness, whilst at the same time deliver performance suitable to the bandwidth increase offered in the longer term by the fiber medium. This paper looks at this issues and debates the opportunities for a new class of optical devices making use of the progress in optical integration

Present-day data networks are being challenged by the relentless growth of bandwidth demand. Most of this future growth, however, is not arising from more users with the same application bandwidth demands, but rather from the dramatic increase in bandwidth demand of emerging applications. The compounding of these two trends has set the stage for optical networking technology to make significant contributions in next-generation data network architecture. In this talk, we will explore key research problems of optical network architecture (such as flow switching) that, if properly exploited, will enable significant growth and cost-effective scalability of next-generation data networks.

A deterministic first-order and second-order polarization-mode-dispersion (PMD) emulator or compensator using binary polarization switches is demonstrated. The rotation angles of polarization switches are ±22.5° and are used in pairs between birefringent elements to provide 0, 45 and 90-degree polarization coupling to generate either first-order (0 and 90-degree) or second-order (45-degree) PMD. We also verify the repeatability of such module for practical applications.

Using the split-step Fourier method, the nonlinear effects and higher-order dispersion in optical fiber are studied and numerically analyzed. Based on the analysis, the theoretical model of higher-order dispersion compensation with phased modulator is presented, with emphasis on the third-order dispersion. According to the theoretical model, simulation models based on VPItransmission is designed. Finally, experimental schemes are designed, and experiment is finished: 200-fs pulses propagate through a 47-km fiber link, including a 39.8-km SMF (single-mode fiber) and a 6.71-km DCF (dispersion-compensation fiber). In that experiment, the oscillating tails are completely suppressed and the third-order dispersion is successfully compensated with the phase modulator. Moreover, the parameters of modulator are optimized with VPItransmission Modeling.

We study the polarization-mode dispersion (PMD) compensation and mitigation performance, using a single PMD compensator to compensate several channels at the same time with degree of polarization (DOP) as the feedback signal in a 2 40 / Gb s × WDM system. Particle-swarm optimization (PSO) algorithm is used to search and track the dynamic degree of polarization (DOP) to achieve the adaptive PMD compensation. New modulation formats are used to mitigate PMD, and the dynamic combination of PMD mitigation taking CSRZDPSK and CSRZDQPSK as the advanced modulation formats have been successfully completed. The numerical results show that the differential phase-shift keyed (DPSK) and differential quadrature phase shift keying (DQPSK) modulation formats are more suitable for PMD mitigation system than the on-off keying (OOK) modulation format. Furthermore, DQPSK format can simultaneously double WDM spectrum-packing efficiency and upgrade the system capacity while it can still has the same mitigation performance as the DPSK format in the PMD compensation system. These features make them greatly attractive for high-speed long-haul WDM system.

The effect of PMD on the three axes of the degree of the polarization (DOP) ellipsoid is discussed in the different modulation format systems. The results show that the minimum axis of DOP ellipsoids is suitable for a feedback signal in the compensation systems. In addition, the mitigation performances of DPSK formats are superior to that of OOK format.

We firstly propose an approach to compensate the polarization mode dispersion (PMD) effectively in optical true time delay (TTD) system composed of a Faraday Rotation Mirror (FRM), a normal 3-port circular. The TTD system can preserve an arbitrary input state of polarization regardless of any external perturbations, and is complete free from polarization effects. We demonstrate experimentally the polarization-stability of the TTD system and discuss the RF correspond of the TTD system.

We demonstrate a tunable chromatic dispersion (CD) compensation technique using a semiconductor optical amplifier (SOA) and a coil of dispersion compensation fiber (DCF). Based on cross-phase modulation (XPM) in the SOA, the transient chirp of the received signal can be adjusted by tuning the drive current of the SOA and the power of clock pulse. In this way, a 10-Gbit/s tunable CD compensation setup, ranging from -40ps/nm to 60ps/nm, is realized without changing the length of the DCF.

We investigate the pattern-dependent decoding failures that occur with electronic dispersion compensation (EDC) when using full optical-field reconstruction, and find that the performance of such an EDC receiver may be degraded by an isolated '1' bit surrounded by long strings of consecutive '0's. By reducing the probability of occurrence of this kind of isolated '1', we experimentally achieved 10Gbit/s on-off keyed signal transmission over 372km field-installed singlemode fiber without optical dispersion compensation.

Optical communications is undergoing a digital revolution, as digital signal processing (DSP) emerges as a practical solution for robust long-haul transmission. In contrast to previous systems, in which optical dispersion compensation was considered a necessity, recently uncompensated long-haul transmission has been demonstrated using DSP at the transmitter, receiver, or a combination of the two. We review the historic developments of electronic signal processing as applied to optical communications before focusing the latest developments, namely digital coherent systems. We then discuss the salient features of the electronic signal processing schemes currently under investigation before discussing the challenges and opportunities offered by electronic signal processing.

We present a simplified model for a simple estimation of the eye-closure penalty for amplitude noise-degraded signals. Using a typical 40-Gbit/s return-to-zero amplitude-shift-keying transmission, we demonstrate agreement between the model predictions and the results obtained from the conventional numerical estimation method over several thousand kilometers.

Nanoscale CMOS technologies are posing new network on chip concepts to IC designers. However, the electronic network on chip design faces many problems like energy consumption, long delay and limited bandwidth. Hence, optical network on chip appears as a good candidate to solve these problems. The advances in nanophotonic technology make it more realistic. A new sparse mesh is proposed for optical network on chip. Two types of non-blocking optical node architecture are also proposed to build up core node and switch node. The new architecture fully utilizes the property of XY routing in 2D mesh network, thus saving the number of microring resonators used. The comparisons are made with traditional mesh in number of microring resonators, loss and energy. The results show that the proposed sparse mesh achieves the best in all the aspects. For example, it uses 68% less number of resonators than the traditional mesh. We simulated 2D sparse mesh optical network on chip, and showed network performance under different traffic loads and data sizes. The results show sparse mesh achieves lower average delay and higher throughput than the traditional mesh.

We present the fabrication of high Q factor micro-ring resonators on SOI substrate by directly focused-ion-beam (FIB) milling. Micro-ring resonators with diameters of 10 μm and 80 μm are fabricated and their corresponding intrinsic Q factors are 4,000 and 130,000, respectively.

Slow light using Stimulated-Brillouin-Scattering (SBS) in optical fiber is now an enabling technology for all-optical network functionalities. A brief review of recent advances in related research will be given, highlighted by optical communication system demonstrations.

Semiconductor optical amplifiers(SOAs) application are not as widespread as one would expect. The reason lies in the severe interchannel crosstalk experienced by the signals in a wavelength-division multiplexing(WDM) link using SOA. Theoretical study on channel crosstalk due to gain saturation in SOAs is presented. A significantly higher penalty is paid at more channels because of the gain saturation characteristic of SOAs. Crosstalk mitigation techniques in SOAs are studied. We have demonstrated theoretically and experimentally that a CW light propagating with signal can mitigate crosstalk in SOAs at channel rates of 40Gb/s. Reduction of crosstalk in SOA by amplifying dispersed WDM signals is confirmed experimentally. Based on the results of the experiment, a novel dispersion management scheme is proposed to prevent the interchannel crosstalk caused by cross-gain modulation.

To improve the reliability and scalability that are very important for large-scale all optical fiber sensor networks (AOFSN), three-level hierarchical sensor network architectures are proposed. The first two levels consist of active interrogation and RNs, respectively. The third level called sensor subnet (SSN) consists of passive FBGs and a few switches. As AOFSN is mainly multiplexed by wired and passive FBGs, the routing algorithm for scanning sensors is determined by the virtual topology of SSN due to the passivity. Therefore, the research concentrates on the construction of SSN and aims at proposing regular and unicursal virtual topology to realize reliable and scalable routing schemes. Two regular types of SSNs are proposed. Each type consists of several sensor cells (SC), square-based SC (SSC) or pentagon-based SC (PSC) and is scaled several times from the SCs. The virtual topologies maintain the self-similar square- or pentagon-like architecture so as to gain simple routing. Finally, the switch architecture of RN is proposed for the reliability of the first two levels; and then, the reliability and scalability of SSN are discussed in view of how much link failures can be tolerant, and how each SC is scaled to maintain the self-similarity, respectively.

In order to integrate local ultra-wideband (UWB) environment into fixed wired networks or wireless wide-area infrastructures and eliminate the high cost of microwave electrical circuits or devices, UWB-over-fiber systems have emerged to exploit the advantages offered by optical fiber. We experimentally demonstrate a photonic UWB monocycle pulse generation and pulse position modulation (PPM). The UWB monocycle pulse with inverse polarity can be generated based on the two differently-biased Mach-Zehnder modulators, subsequently the PPM of the UWB monocycle pulse is realized through the nonlinear polarization rotation (NPR) arising in a semiconductor optical amplifier (SOA). We proposed and experimentally demonstrated an effective procedure to generate and maintain a linear output state of polarization (SOP) during the course of the nonlinear polarization rotation (NPR) of an SOA. High performance of polarization switching can be achieved. The proposed scheme is more preferable for the UWB-over-fiber applications due to its compact size and high stability.

A time- and wavelength-interleaved (T- W-interleaved) optical pulse source based on optoelectronic oscillator is presented. By utilizing spectral slicing, a multi-loop configuration is formed in OEO system, which can effectively suppress side-modes. In addition, this configuration is able to generate multi-wavelength optical pluses which can be arranged T- W-interleaved optical pulse stream at different nodes. In this system, 10-ps-wide optical pulses at 5GHz repetition rate are generated with 1ps timing jitter (in the 100Hz-1MHz range) and -100dBc/Hz spurious modes suppression, Along with the optical outputs, 5GHz electrical signal with low phase noise (-113dBc/Hz at 10kHz offset from the carrier) is also obtained.

We demonstrate two all-optical methods for UWB pulse generation based on various nonlinearities of the semiconductor optical amplifier (SOA), namely, self phase modulation (SPM), and cross gain modulation (XGM). In the first method, we present UWB doublet generation based on SPM. The monocycle pulse is generated from dark return-to-zero (RZ) signal, and then converted to doublet pulse by injecting an additional probe signal with the SMF transmission. For the first time to best of our knowledge, we report that the generated doublet pulses are transmitted over 5km SMF by proper dispersion compensation without distortion. Second, we present UWB doublet generation by XGM of two cascaded SOAs. The configuration of our all-optical methods is compact and simple.

It has been investigated of nonlinear propagation of femosecond pulse and supercontinuum generation (SCG) in three photonic crystal fibers (PCFs) with different dispersion profile in the 1550nm window by numerical stimulation. The influence of higher-order effects on supercontinuum, including higher-order dispersion (HOD), self-steepening (SS), and stimulated Raman scattering (SRS), are discussed in detail as well. The results of numerical simulations show that group velocity dispersion and self-phase modulation take main effect in the initial stage. In the PCF with anomalous dispersion, SRS plays main role when the propagation distance increases, which induces a red-shift of the central wavelength, suggesting the appearance of soliton self-frequency shift. In the PCF with near-zero anomalous dispersion, HOD plays main role. This results in the fission of higher-order solitons and remarkable broadening of the pulse spectrum. In the PCF with normal dispersion, higher-order effects have almost no effect on the pulse. The Gaussian pulse wave broadens to rectangular symmetrically, and the pulse spectrum broadens symmetrically, too. However, the broadening is smaller than the former two cases.

A time- and wavelength-interleaved optical pulse source, which is based on spectral slicing and dispersion management, is proposed and experimentally demonstrated. The pulse source has a simple configuration, and both the repetition rate and the wavelength channels can be tuned easily.

Using POLSK (POLarization Shift Keying) and DPSK (Differential Phase Shift Keying) formats we fully exploit Semiconductor Optical Amplifiers (SOAs) for signal amplification in practical WDM Nx10 Gbit/s systems. Record transmission experiments are presented, thus making SOAs a competitive amplification alternative in metro and extended-metro networks.

Optical fiber transport networks have been evolving rapidly to meet the demands of today's telecommunications by providing unprecedented transmission capacity and reach. In dense wavelength-division-multiplexing (DWDM) transport systems, 10-Gb/s channels are widely deployed, and 40-Gb/s channels are starting to be added in the same systems. In the foreseeable future, 100-Gb/s channels are expected to be carried. The realization of such high spectral- efficiency DWDM systems with mixed data rates and signal formats presents several technical challenges. In this paper, we review these challenges and discuss promising technologies that may potentially address these challenges.

Digital signal processing (DSP) based coherent detection attracts extensive studies recently. DSP technology simplifies coherent detection and brings it more advantages and channel impairments can be compensated electrically. The lumped dispersion compensation scheme makes the fiber link simple, flexible and potentially cost effective. However, such pseudo-linear long haul transmission systems are greatly degraded by intra-channel nonlinearities at high bit rates. The optical pulses are highly dispersed and overlap strongly during propagation, resulting serious intrachannel four-wave mixing (IFWM). IFWM introduces pattern dependent nonlinear phase shifts to the signals. It has been pointed out that IFWM induced phase noise are correlated from symbol to symbol. The autocorrelation function has been studied in detail. Here we extend the result to fiber link with single mode fiber (SMF) for transmission only. We studied the statistics of IFWM induced phase noise in Return to Zero-Differetnitally coded Quadrature Phase Shift Keying (RZDQPSK) transmission systems with identical fiber spans and SMF fiber spans numerically. The results show that the phase noise correlation of neighbouring bits is reduced for the latter one but the correlation length is increased with fiber span number and chromatic dispersion value.

A method for subcarrier demultiplexing in Radio-over-Fiber system has been proposed and verified experimentally. A continuously tunable single resonance microwave photonic filter using the tunable fiber Mach-Zehnder interferometer (MZI) as the slicing filter, is implemented to realize the demultiplexing. Two subcarriers, one is 2.4GHz representing the wireless LAN signal , the other is 900MHz representing the Cellular signal, can be demultiplexed separately. The proposed approach has been demonstrated both theoretically and experimentally.

Hybrid fiber-wireless networks for fixed wireless access operating in the millimeter-wave wave (mm-wave) frequency region have been actively pursued to provide untethered connectivity to ultra-high bandwidth communications. The architecture of such radio networks requires a large number of antenna base stations with high throughput to be deployed to maximize the geographical coverage with main switching and routing functionalities located in a centralized location. The transportation of mm-wave wireless signals within the hybrid network is subject to low opto-electronic conversion efficiency, fiber chromatic dispersion and also signal degradation due to nonlinearity along the link. One of the major technical challenges in implementing such networks lies in the mitigation of the various optical impairments that the wireless signals experience within the hybrid network. In this paper, we present an overview of the different techniques and schemes to overcome some of the signal impairments in transporting mm-wave radio signals over optical networks.

White LED (light emitting diode) offer advantageous such as high brightness, lower power consuption and long life spand, and is expected to be the next generation of lamp. The other merits of LED are high sensitivity and fine modulation performance. An indoor optical wireless communication system is propsed and three kind of modulation schemes are compared in band requirement and power requirement. Nine LED lights are installed in the different of the room celling, so the RMS (root-mean-square delay spread) delay is also consided in hight bit rate enviroment, the power loss and BER are presented under different time delay. The simulation results show that, the MPPM is bandwidth efficeient than PPM and OOK. At the condition of high bit rate, MPPM performance much better than OOK and PPM with low BER.

Experimentally, investigation of nonlinear polarization evolution induced by cross phase modulation in optical fiber and its impact on transmission link are circumstantiated in this letter. We experimentally demonstrate that the use of 500-m length of highly non-linear fiber readily allows for our novel approach of implementing high-speed format transformation from return-to-zero amplitude-shift keying (RZ-ASK) to return-to-zero frequency-shifting keying (RZ-FSK) at 40Gb/s and above based on cross-phase-modulation-induced polarization rotation. The high-performance feasibility is verified and the operating parameters are achieved.

In this paper, we propose and experimentally demonstrate an all-optical non-return-to-zero (NRZ) to return-to-zero (RZ) format conversion using semiconductor ring laser (SRL). A 155Mb/s NRZ optical data signal is injected into the counterclockwise (CCW) direction of the SRL and an optical logic clock signal is injected to the clockwise (CW) propagation direction. The power and wavelength on both sides are adjusted so that during the 'high' level of the clock, the SRL is held to the CW direction regardless of the logic value of the corresponding data bit. The SRL will only be switched to the CCW direction when the input data is 'high' and when the clock level is 'low'. Therefore the NRZ format signal is converted to RZ format signal, with a duty cycle decided by the duty cycle of the optical clock. Extinction ratio of > 10 dB has been achieved and the scheme can also be used for all-optical 3R. Although the speed of the signal is yet limited, higher speed can be achieved with faster SRL devices.

We propose and demonstrate all-optical format conversion from nonreturn-to-zero differential phase-shift-keying (NRZ-DPSK) to return-to-zero binary PSK (RZ-BPSK) at different bit-rates, using a tunable band-pass filter, a semiconductor optical amplifier (SOA) fiber ring laser and an SOA based Mach-Zehnder interferometer (SOA-MZI). The filter acts as the DPSK demodulator and also as the differential decoder. It converts the NRZ-DPSK into the RZ signals, which is equivalent to the data information before differential encoding. The SOA fiber ring laser is used to recover the clock signals from the demodulated RZ signals at different bit-rates. In the subsequent SOA-MZI, the RZ signal is used to modulate the recovered clock signal which is synchronous to the demodulated RZ signal. Thus, the amplitude information is encoded onto the phase of the clock signal, through cross gain modulation (XGM) and cross phase modulation (XPM) effects in the SOA-MZI. The converter can be operated at flexible bit-rates and used as the interface between long-haul WDM and OTDM systems.

Based on a Mach-Zehnder interferometer and optical temporal time delay (OTDL), we propose and demonstrate experimentally an all-optical 40-Gb/s RZ-DPSK to ASK-Manchester format conversion. This obtained ASK-Manchester signals have the phase changes due to the demodulation of RZ-DPSK, which is called as phase-incorporated ASK-Manchester signal. Compared with pure Manchester signal, the simulated results show that the phase-incorporated ASK-Manchester signal has a compact optical spectrum and lower power at the low radio frequency (RF) frequencies, which presents high tolerance on chromatic dispersion and little the cross-talk between payload and label. Moreover, the efficiency and the BER performance of conversion and the dynamic range of fiber input power, transmission length, and chromatic dispersion tolerance of the converted ASK-Manchester are detailed discussed. Less than 1-dB power penalty over 50-km SMF transmission is observed.

A novel digital-to-analog converter based on multi-wavelength pulse source is proposed and primarily demonstrated by experiment. The multi-wavelength pulse train which has been pretreated is separated in time domain through dispersion. The pulse of different wavelength corresponds to the input digital data. Only one electro-optical modulator is needed in the scheme. In the primary experiment, a 3-bit photonic digital-to-analog converter with a sample rate of 2.5 Gb/s is realized. The scheme is simple and compact compared with the scheme using multiple electro-optical modulators.

A novel high speed Optical Frequency-domain Transmission method and System (OFTS) based on two time lenses is proposed and analyzed in this paper. By using the unchanged spectral profile through light propagations in fiber, the long-haul optical transmission over 100Gbps per channel can be achieved with highly tolerance of linear distortion such as PMD, chromatic dispersion, time jitter etc. In the novel method and system, a time lens is used at the transmitter to convert the initial time domain signal into frequency domain signal. And at the receiver, another time lens is used to transform the received frequency domain signal back to time domain. To prove the advantages of OFTS, a specifically designed 10×100Gbps OFTS is numerically simulated over 1880km transmission without PMD compensation.

Key issues related to the modeling of ultra-high speed transmission are discussed in this paper. These topics include components modeling, link modeling and BER estimation. Different solutions for the transport of 100 Gb/s over a single wavelength including technologies such as coherent detection, polarization multiplexing, optical OFDM, and digital signal processing are reviewed and compared with means of numerical simulations.

The rapidly increasing bandwidth demand requires deployment of 40 Gbit/s systems and 100 Gigabit Ethernet in the near future. Different reach applications exploit advanced modulation formats employing spectral shaping and phase coding. The resulting different solutions for photoreceivers are summarized and efforts within the industry towards integrated solutions are described. First results of integrated DPSK receivers with co-packaged delay interferometers will be presented.

We propose a method to generate high-speed frequency-shifted keying (FSK) signal by demodulating two phase-shifted keying (PSK) signals with proper wavelength detuning. The generated FSK is easy for detection and has a small crosstalk to optical label signal. In an 85km SMF transmission demonstration, we achieve 1.5dB power penalty in 10Gbit/s FSK payload and 1dB power penalty in 155Mbit/s amplitude-shifted keying (ASK) label. Compared with the phase continuous FSK, this kind of FSK occupies wider bandwidth due to its phase discontinuity but is simpler in its generation.

Polarization-mode dispersion compensation performance of optical DQPSK is studied in this paper. Firstly, the simulation model for the back to back system of optical DQPSK is established by using Matlab. The optical power spectra and eye diagram of different DQPSK modulation formats are obtained. Then, we adopt two pseudo random binary sequences as input data streams. In the back to back DQPSK system, by means of precoding, modulation and demodulation we obtain exactly the same as the original signal. The system for simulation is verified. At last, we add the polarization-mode dispersion compensation program module to the back to back DQPSK system. The result is shown that the modulation formats with smaller bandwidth have better PMD compensation performance. and that PMD compensation performance of DQPSK is better than that of OOK and DPSK. CSRZDQPSK combines the carrier-suppressed feature of CSRZ formats and specific spectrum characteristic of DQPSK formats, thereby it has the best performance of PMD compensation performance.

A method called coherent hybrid is proposed to monitor the deficiency of differential phase shifted keying (DPSK) either generated by the bias offset drift of Mach-Zehnder modulator or an unmatched drive voltage applied to a phase modulator (PM) is proposed and demonstrated. The hybrid power ratio of the maximum to the minimum after mixing the unmodulated local oscillation lightwave with the modulated signal from the same light source is employed to measure the degree of deficiency, simultaneously eliminate the dependence on light source power. Our experimental results reveal that a±5% bias offset of MZI can produce 0.3dB power variation, larger than that of 0.02dB in the traditional method by measuring the signal average power, and a drive voltage with a more than 10 percent deviation from the halfwave voltage can be detected when it is applied to a PM. A feedback and control circuit based on this scheme is developed and applied in our system. The impairments given by the unparallel orientation between the polarizations of signal and local oscillation lightwave and the transient time of modulation signal pulse is discussed. By properly setting the control parameters in the feedback control process, the impairments can be minimized.

All-optical performance monitoring based on optical sampling was proposed for multi-bitrate and multi-format signals in transparent optical networks. The performance of 42.7-Gb/s RZ-DQPSK and 10.7-Gb/s RZ-DPSK signals were successfully monitored and correlated with the actually measured BER.

Progress in optical network has stimulated interest in optical performance monitoring (OPM). Optical sampling technique is believed to be a promising candidate to monitor the physical state of the network. The theoretical model of a monitoring system based on optical sampling in semiconductor optical amplifier (SOA) and software synchronized algorithm is constructed. Compared with the results obtained by Optsim, the monitoring system model is proved. For 10Gb/s NRZ (RZ) optical signals, the differences on Q values between the results obtained by SOA and the ideal sampling processes are 0.195dB (0.247dB), 0.988dB (0.594dB), and 1.707dB (0.596dB) for pump energy equal to, ten times of and fifty times of the optical data signal energy respectively. The sampling device can induce degradations of the sampling results. It is mainly because of the gain saturation and the nonlinear effects in SOA. The high input power can make the gain saturated deeply and further influence the probe and conjugate outputs. At the same time, the pump power should not be too low. The proper pump power will obtain better sampling linearity and better sampling results.

We present a novel method to generate ultra-wideband (UWB) monocycle pulses using polarization state rotation filtering of optical negative pulses. Monocycle-like optical pulses with reversed polarities can be generated with appropriate polarization adjustment.

The supercontinuum (SC) spectrum generation is proposed in dispersion-flatted fibers with concave dispersion profile (DFF-CCDP). The effects of pulse characteristics and the fiber parameters on the SC spectrum characteristics are numerically investigated in the DFF-CCDP by using the split-step Fourier method. The characteristics of the SC spectrum become worse when the parameters are much smaller or greater. The parameters should be set according to the practical applications.

Effects of chirp and noises on autocorrelation characteristics of the linear chirped hyperbolic secant pulse are numerically investigated. It is found that the intensity autocorrelation curve and the pulse waveform are unchanged with the increase of the chirp parameter |C|. The spectral width and time-band product of autocorrelation spectral curve increase with the increase of |C|. The autocorrelation curves broaden near the center of the curves and heighten at the edges with the increase of the random noise, temporal window and sampling number. A method of filtering the random noise is given which is useful for us to use the SHG-FROG analyzer or the instruments based on autocorrelation technology.

An enhanced theoretical model for wavelength conversion of SOA is presented. The chirp for the converted signal is analyzed and confirmed by numerical simulation.

Considering that shot noise is Poisson distribution and the envelope of background optical noise is Rayleigh distribution, the Gaussian Approximation method is firstly used to evaluate the performance of Satellites Laser Communications precisely. The Moment Generating Function expression of considering shot noise, background optical noise and hot noise is firstly derived. The system performance of On Off Key modulation of traditional method is compared with GA method in direct detection system. Results indicate that the performance of Satellites Laser Communications by GA is different from traditional method.

In this paper, we theoretically and experimentally introduce a new reconfigurable optical add/drop multiplexers (ROADM) module which bases on the narrow-band optical switch (NBOS) technology and achieves the narrow-band optical switching with a Fabry-Perot cavity. In comparison with the analysis of the current ROADM technologies, the NBOS-based ROADM module is proven to be cost-effective, low insertion loss, narrow-band, and has a more mature technology. It would grow to be a cost-effective ROADM solution than any of the others.

The bias drift effect in the packaged LiNbO₃ modulator is investigated. The Bessel expansion of the dithered clock shows that the harmonic component equal to the dither frequency can be synchronously demodulated to get the bias drift and avoid the random phase difference between the clock or data and the dither signal. By using the time division control method one control system can track two modulator bias drift in 40Gb/s RZ optical transmission system because the optimum bias point changes very slowly.

A 2×2 and a 1×4 fiber-type optical phased array with PZT phase shifters have been implemented. Fibers are adopted as transmission paths and emitting antennas in the optical phased array system. By using the feedback optimization we had proposed before, the inaccuracy of fiber length is overcome. By adjusting the phase shift of each PZT phase shifter repeatedly according to the algorithm we had proposed for optimization, 1-D and 2-D beam forming and beam steering in a short period of time have been observed respectively. Moreover, a 1×6 fiber array is also tried. Although the interference is not stable enough, much narrower beams are achieved.

This paper describes the design of a burst-mode ultra-long haul PON trunk fiber protector based on SOA. It's transparent to protocols and thus compatible with various PON systems. With this technology, the robustness and transmission distance of PON network can be greatly improved and extended. The fiber protector supports a maximum transmission distance of 50km and a 1:128 branching ratio. It supports automatic protection switching of trunk fiber whose longest protection switching time is less than 20ms. In addition, in the design of the ultra-long haul PON system, by EPON as an example, the ranging of EPON system, RTT compensation and other parameters also need to be amended. As a result, this article also makes research and description on synchronization, ranging, time delay compensation, and other key parameters of the EPON system.

The effect of non-ideal dispersion and reflection characteristics of chirped fiber Bragg gratings (CFBG) on the 40-Gb/s and 10-Gb/s transmission is investigated. The effect of group delay ripple (GDR) on eye-opening penalty (EOP) of carrier-suppressed return-to-zero (CSRZ) and return-to-zero (RZ) formats is analyzed and compared. The system penalty for different amplitude and period ripples are quantified.

Asymmetric lithium niobate Mach-Zehnder interferometer and its applications in photonic analog-to-digital conversion will be discussed. Two schemes based on the asymmetric interferometer will be proposed and analyzed. The first scheme is the phase shift photonic analog-to-digital conversion using asymmetric interferometer and synchronized multiwavelength optical sampling pulses. Because of the dispersion effect of the lithium niobate crystal, when multiwavelength optical pulses enter into the interferometer, at the output port, different wavelengths will have different phase differences between two arms. As a result, after interference, the transmission characteristics of different wavelengths will have a phase shift between each other, and this is just the key issue of phase shift photonic analog-to-digital conversion. The other scheme we will propose in this paper is a spectral encoded photonic analog-to-digital conversion. The spectral transmission characteristic of the asymmetric interferometer will shift with the voltage change of the analog signal, and this shift has an ideal linear relation with the analog voltage change. The peak wavelength of the transmission spectrum can be detected to realize quantization of the applied analog signal. Using both schemes presented in this paper, high sampling rate and high resolution optical analog-to-digital conversion can be realized.

A solution is proposed, which combined with FSO communication and laser ranging in order to effectively integrate two payloads. It is a transformation of feedback pulsed laser ranging. Laser ranging hardware and FSO communication hardware share a lot in common. Detailed description of working principle, error analysis and elementary experiment result is also included. The further analysis indicates that the measurement precision can reach sub-millimeter by counting more ranging period. Moreover, when temperature varies 30°C, the measure distance varies 2.25cm. In this case, temperature control is necessary for sub-millimeter measurement.

Due to the Multipoint-to-Point nature of the uplink, the upstream data transmission in a GPON system is burst-mode, and both the guard time and preamble time are short. This burst-mode nature of the GPON uplink brings many challenges for the design of the burst-mode receiver. This paper presents a newly designed AC-coupled burst-mode optical receiver. The AC-coupled receiver uses a large time constant and a high-speed reset switch. The concept of cumulative bit difference of pattern-dependent jitter is also put forwarded in this paper. Finally, simulation results are provided to show the feasibility of the scheme.

Conversion from a 10 Gb/s non-return-to-zero (NRZ) signal at 1550 nm to a 10 Gb/s return-to-zero (RZ) signal at 1561 nm has been experimentally verified based on four-wave mixing (FWM) in a highly nonlinear photonic crystal fiber (HNL-PCF) without the need of an extra NRZ-to-RZ conversion step. The signal light at 1550nm was modulated by a NRZ electric pulse pattern. The pump light at 1555.540 nm generated by a mode locked laser (MLL) was modulated by a clock signal. When these two lights were controlled strictly and amplified to about 22dBm, the FWM effect happens in HNL-PCF. Two new spectrum peaks were achieved at 1545 nm and 1561 nm respectively. Then the peak at 1561nm was filtered out and discussed. Then we changed the wavelength of signal light from 1543nm to 1569nm. The FWM effect appeared in this whole wavelength range. But the best quality of conversion signal was achieved at 1550nm. So when the signal at 1561nm was filtered out , it was a pulse signal carrying information. The signals at other wavelength were pulse signals without information.

A novel method to generate ultra-wideband (UWB) signals based on two differential group delay modules and one Mach-Zehnder Modulator is proposed. Both simulation and experiments verified that the method could generate two kinds of UWB monocycles using low bit-rate non-return-to-zero data source (1Gbps). The monocycles have a fractional bandwidth of 127% at centre frequency 5.5GHz.

To achieve adaptive modulation technology in Optical Orthogonal Frequency Division Multiplexing (OOFDM) system, the channel estimation has to be made first and foremost. By imposing a symbol of pilot signals after a certain amount symbols, the frequency response of a given transmission link can be given out. The modulation format can be determined according to the frequency response. A high (low) modulation format is used on a sub-carrier suffering a low (high) transmission loss. The modulation format can be adjusted adaptively.

A numerical solution, travelling-wave-like spatial soliton is obtained by using travelling-wave transform along with numerical methods to solve the photorefractive nonlinear equation, and its necessary forming conditions are presented. The evolution properties of the travelling-wave-like spatial soliton in the photorefractive crystal are investigated. The results show that travelling-wave-like spatial soliton can not only keep its shape invariably, but also translate its energy transversely. It is possible for controllable deflection of the spatial soliton. The deflection angle depends on the magnitude of the transverse phase modulation coefficient. These properties of the multi-soliton coherent interactions can be potentially applied in optical computation.

The transmission characteristics of a frequency shift-keying signal at 40Gb/s which generated by a novel scheme are investigated under varying dispersion management. The resilience of fiber span, compensation ratio, and power level are obtained.

In this paper, a novel method to generate both monocycle and doublet UWB pulses is demonstrated, where pulse shape modulation(PSM) can be easily implemented. Only two wavelengths and two modulators (one dual-in dual-out modulator) are applied to achieve PSM. The data driving the first modulator is set to be 250Mbit/s 107-1 pseudo-random bit sequence (PRBS). The 1GHz pulse pattern is synchronised with the data. The electrical spectrum of the signals processes the centre frequency of 4GHz and -10dB bandwidth of 5.9GHz. The fractional bandwidth is about 147.5%, which matches the FCC standard.

In this paper, a scheme for all-optical AND gate which exploits the cross-phase modulation (XPM) effect in an asymmetric semiconductor-optical-amplifier-assisted Mach-Zehnder Interferometer (SOA-MZI) has been proposed for the first time. No additional input beam such as a clock signal or continuous light is employed in the design, which is required in other AND operations. The scheme is validated and the system performance under various parameters is investigated through numerical simulations. This logic gate is helpful for future all-optical signal processing configurations.

A novel optical receiver in wireless optical communication system is presented. The receiving method is demonstrated based on an optical taper. The taper is mounted near the focus spot of the receiving lens along the principal axis. The larger-area end of the taper receives the converging beam and the smaller-area end transmits optical power to detector or optical fiber. In this way, the immunity to light beam deviation in wireless optical communication system can be enhanced evidently. The theoretical and experimental results show that the system has a perfect performance and realizes a low loss of 4dB for the range of field of view within 1mrad.

Self- and cross-phase modulation (SPM, XPM) in fiber is expected to play a major role in ultrafast all-optical analog-todigital conversion(ADC), which can overcome the obstacles of inherently limited operating speed of electronic ADC. In this paper, we propose an optical quantizing and coding method for all-optical ADC based on asymmetrical nonlinear optical-loop mirrors . The multiperiod transfer function, which is the key to quantizing and coding, is achieved through a careful design of the asymmetrical nonlinear optical-loop mirrors. We conducted proof-of-principle numerical simulation and successfully demonstrated 3-bits all optical ADC with Gray code output, it is ready for experiments in future. The proposed optical quantizing and coding, combined with existing optical sampling techniques, will enable ultrafast photonic ADC without electronics.

A fast and accurate polarization control and stabilization system, which is based on coordinate system transformation, is experimentally demonstrated. The typical control speed is 50-ms and the output SOP precision is 0.5°.

A novel all-optical frequency up-conversion method implementing microwave photonic filter in a radio-over-fiber (RoF) link is proposed and demonstrated. The electrooptical phase modulator, electrooptic intensity modulator in combination with 75-km single-mode fiber (SMF) is used to form the microwave photonic filter and fulfill the frequency upconversion function at the same time. The technique requires large modulation index at the phase modulator, which simultaneously increases the efficiency of the frequency conversion. The proposed approach has been demonstrated both theoretically and experimentally.

Recent years the application of free-space laser communication system is become more important.A smart antenna system combines multiple antenna elements with a signalprocessing capability to optimize its radiation and/or reception pattern automatically in response to the signal environment. In this paper, we mainly study the Optical adaptive Antenna Array in the free-space laser communication.

Multimode VCSEL(Vertical Cavity Surface Emitting Laser) coupling into MMF (Multimode Optical Fiber) and its influence on MMF link has been numerically studied. Considering the transverse misalignment between VCSEL and MMF, coupling efficiency between VCSEL higher-order modes and modes of MMF is calculated and the Mode Power Distribution in the MMF is obtained. Simulation of MMF link with multimode VCSEL indicates that the higher-order VCSEL modes lead to decreasing of the eye opening of received signal. The transverse misalignment of VCSEL between MMF greatly deteriorates the eye opening. With 300m OM3-MMF link , the EOP (eye-opening-penalty) is greater than 14dB when the misalignment between VCSEL and MMF is lager than 5μm.

With service competitions of carriers aggravating and client's higher service experience requirement, it urges the MAN technology develops forward. When the Core Layer and Distribution Layer technology are mature, all kinds of reliability technologies of MAN access Layer are proposed. EoRPR is one of reliability technologies for MAN access network service protection. This paper elaborates Ethernet over RPR technology's many advantages through analyzing basic principle, address learning and key technologies of Ethernet over RPR. EpRPR has quicker replacing speed, plug and play, stronger QoS ability, convenient service deployment, band fairly sharing, and so on. At the same time the paper proposed solution of Ethernet over RPR in MAN, NGN network and enterprise Private network. So, among many technologies of MAN access network, EoRPR technology has higher reliability and manageable and highly effectiveness and lower costive of Ethernet. It is not only suitable for enterprise interconnection, BTV and NGN access services and so on, but also can meet the requirement of carriers' reducing CAPEX and OPEX's and increase the rate of investment.

We focus the effect of stimulated-Raman-scattering (SRS) on modulation instability (MI) in anomalous-dispersion regime of optical fiber from extended nonlinear Schrödinger equation. An expression of the gain of MI is given and the ranges and the growth of the gain spectrum are theoretically and numerically analyzed under considering SRS or neglecting SRS. The gain peak is also theoretically demonstrated. The effect of SRS on the frequency spectrum of MI is few when incident power is small. And incident power increases, the spectrum varies obviously, the range and the growth of the gain is lower if considering SRS than that of the gain if neglecting SRS. We also find that the maximum incident power will be limited by the fourth-order dispersion and will be enlarged by SRS to an extent.

In this paper, a partial inner wavelength method (PIWM) is proposed to enlarge buffering capacity of shared FDLbuffers. An optical packet switch called OPS-PIWM is proposed to realize the PIWM. In addition, two optical packet scheduling algorithms, i.e., greedy scheduling algorithm (GSA) and conservative scheduling algorithm (CSA) are proposed to improve the performances of OPS-PIWM. The packet loss rate of OPS-PIWM versus the number of fiber delay lines, the number of tunable wavelength converters, and the number of inner wavelengths are evaluated by simulation experiments.

A statistical model of the aggregation operation in an optical packet based network interface is presented. The operation aggregates client-layer packets into optical slots, and includes a timeout parameter to limit the maximum delay. Aggregation serves the purpose of providing an efficient filling of the large optical slots with smaller packets from the client layer. The model will be utilized to derive an optimal choice of timeout parameter that will lead to an efficient utilization of the optical packets.

This paper discusses electronics dispersion compensation (EDC) by decision feedback equalizer (DFE) in high speed optical communication system using the LMS and the RLS algorithms, respectively. The theory of the decomposed DFE derived from basic concept of DFE in EDC is studied. Compared with the traditional DFE, the system complexity of new one has been reduced while it still keeps the high performance of DFE. Simulation results prove the validity of practicability and indicate that the decomposed DFE using RLS algorithm can more effectively eliminate intersymbol interference (ISI) after being transmitted long distance.

The detailed origin of phase jitter in fiber-optic communication systems was firstly reviewed. An overview of distribution of received phase jitter was presented. By contrasting the phase difference of sample pulses, variance of phase noise of an DPSK system in dispersion-managed finite fiber spans was calculated and simulated by Matlab, and the Gaussian shaped possibility density function (p.d.f.) was obtained. The performance of signal was analyzed by estimating BER, which was calculated from the p.d.f.. By adding ASE noise of different power into the signal at each EDFA, we found that the suppression of ASE noise contributing considerably to the improvement of signal performance.

We compared the theoretical bit-error-rate (BER) and throughput performance limits of 2-D time-wavelength and coherent time-spreading optical code division multiple access (OCDMA) communication systems. The results show that the BER performance of the coherent time spreading OCDMA scheme is much superior to 2-D time wavelength scheme in the multiple access interference (MAI) noise limited case without considering beat noise, and the beat noise affects the system performance seriously and appears as a dominant source of noise for the two kinds of OCDMA systems. Nevertheless, the 2-D time wavelength scheme has better tolerance to beat noise than coherent time spreading OCDMA scheme, and the superiority is more apparent with the increasing of the code length N. While the coherent time spreading scheme has higher peak throughput, and it supports a higher throughput than the 2-D time wavelength scheme at higher offered loads.

An analysis of the code parameters that are important to asynchronous coherent time-spreading OCDMA system is presented. Relationship between the mean intensity of aperiodic cross-correlation and multiple-access interference (MAI) and beat noise (BN) is deduced. Considering the mean system performance, MAI and BN can be calculated directly from aperiodic auto-correlation function, and the aperiodic cross-correlation function is not needed. Hence, a computational saving can be achieved For 511 length Gold sequence, coherent time-spreading OCDMA system can support 12 interfering users for the mean performance.

A new all-optical broadband phase modulator of subcarrier in radio over fiber (RoF) system has been proposed. This modulator is based on cross phase modulation (XPM) in fiber. In this approach the stability of the subcarrier modulation is improved significantly as most of the disturbance introduced by environment variance can be cancelled when the lightwaves interfere in PD. A proof-of-concept experiment demonstrates the feasibility of the modulator.

An optical-domain wideband microwave amplification system which takes advantage of the large bandwidth capacity of optical devices to amplify optically carried microwave signals is proposed. A comprehensive theoretical model of the system including link gain, noise and spur-free dynamic range (SFDR) is established, with which all kinds of performance of the system can be analyzed conveniently for given system parameters. Based on the theoretical model, an experiment platform for the microwave amplification system is established. The source microwave signal is modulated up to the optical domain by carrier-suppressed modulation in a Mach-Zehnder intensity modulator (MZM) whose DC bias voltage is controlled by an automatic-bias-voltage-control (ABVC) module. The output optical signal from the MZM is amplified by a commercial erbium-doped-fiber-amplifier (EDFA) and then passes an optical band-pass filter (OBPF) to filter out out-band amplified spontaneous emission (ASE) noise. Finally the optical signal is fed into a photo-detector (PD) and the amplified microwave signal is obtained at the output of it. Owning to the wide gain spectrum and the large small-signal gain of the EDFA, wideband microwave gain which is larger than 17 dB is obtained over 12GHz-bandwidth. The microwave gain is quite stable while the input microwave power varies from -60 dBm to -15 dBm and the input microwave frequency is tuned from 100 MHz to 12 GHz. The noise floor is approximately -137 dBm /Hz, which is mainly caused by the ASE noise brought in by the EDFA. The SFDR of the system is third-order inter-modulation dominated and is measured to be 113.1dB•Hz^2/3.

A novel integrated demultiplexing photodetector which consists of a filter cavity and an arc absorbing cavity is proposed in this paper. The top mirror of absorbing cavity is designed to be curved, which can increase the times of light reflection. The maximum quantum efficiency of TE lightwave can reach 81.84%, while maximum of TM is 65.62%. An obvious near-to-zero domain of reflectivity of TM lightwave leads to its low quantum efficiency. The impact of incident angle on quantum efficiency is demonstrated by comparing two series of incident angles in two devices which have high and low quantum efficiency respectively. In addition, the spectral response linewidth of this novel photodetector is less than 0.8 nm.

A novel power splitter based on self-imaging phenomenon in multimode heterostructure is designed and analyzed. Such a photonic crystal waveguide is a structure combining square and hexagonal photonic crystal lattices. The size and transmission of our designed new power splitter is much smaller and higher in comparison with the conventional MMI power splitter. The device can be applied to optical communication systems and be integrated easily with other optical devices. The finite-difference time-domain (FDTD) method is adopted for the numerical simulation of related structure. This approach can be extended to novel design of MMI device based on photonic crystals.

As multicast applications become more and more popular, implementing multicast in optical domain has attracted more and more attention due to its more efficient bandwidth usage. The MC-OXC is the key for realizing multicasting at optical layer and directly affects the performance of the multicast services in WDM networks. In this paper, a novel mixed-integrated optical multicast switch for multicast-capable cross-connect is proposed. The new architecture uses power splitters for multicast connections only, allowing unicast sessions to pass without enduring unnecessary power loss. And also the performances of the MC-OXC are analyzed. We performed the cost and cost sensitivity analysis for different MC-OXCs. The results show that the proposed architecture guarantees strictly nonblocking with lower costs.

Optical packet switching (OPS), which transfers the switching function from electrical domain to optical domain and provides the smallest switching granularity, is the most potential candidate of switching form in the future optical networks. Optical code division multiplexing (OCDM) is the mostly practical all-optical processing technology at the state of the art. The experiments of optical packet switching with optical code (OC) label have demonstrated the switching capability and advantages. But the timing of erasing and inserting label, which is similar with the bit-serial label processing, is the stringent requirement of this scheme. OCDM optical packet switching, which encodes the payload directly and removes the label when the payload is recovered at the decoder, has no stringent timing requirement. Multiple access interference (MAI) is the main factor degrading the performance of OCDM optical packet-switched networks. In this paper, the effects of MAI are studied at the end of optical label path where the packets experience multiple hops. For eliminating the end-to-end BER, the optical label paths need to be established in an optimum way and the load are required to be balanced. One load-balancing algorithm based on the end-to-end BER of OCDM path is proposed to improve the network performance.

Optical burst switching (OBS) is regarded as a promising optical switching paradigm for next generation optical Internet. More and more real-time traffic is transported by IP and it is important for OBS to support real-time traffic. Based on the delay analysis of OBS networks, a pipeline scheme is proposed to reduce the delay and efficiently support real-time traffic for OBS networks in the paper. When real-time traffic is transported in the core nodes of the established pipeline and its burst head packet can not need to be processed. So the pipeline scheme reduces the offset time and delay of realtime traffic. The simulation result shows that the pipeline scheme effectively reduces the end-to-end delay of real-time traffic in the OBS networks.

In this paper, the delay issue in unequal outputting probabilities applied optical burst switching (OBS) network has been proposed and analyzed. The average waiting length of queue in optical buffer and the average total length of queue in OBS core router have been given. Analytical results indicate that, the performance of delay will affected by the ratio among different length traffic.

Routing and wavelength assignment (RWA) is one of the key issues in the wavelength-routed optical networks. Although some RWA algorithms have been well performed to meet the need of certain networks requirement, they usually neglect the performance of the whole networks, especially the load balancing of the whole networks. This is quite likely to lead to some links bearing excessive ligthpaths and traffic load, while other links being at an idle state. In this paper, the load distribution vector ( LDV ) is introduced to describe the links load of the networks firstly. Then by means of minimizing the LDV of the networks, the load balancing of the whole networks is tried to improve. Based on this, a heuristic load balancing (HLB) strategy is presented. Moreover, a novel RWA algorithm adopting the heuristic load balancing strategy is developed, as well as two other RWA algorithms adopting other load balancing strategies. At last, the simulations of the three RWA algorithms with different load balancing strategies are done for comparison on the basis of both the regular topology and the irregular topology networks. The simulation results show that the key performance parameters such as the average variance of links, the maximum link load and the number of established lightpath are improved by means of our novel RWA algorithm with the heuristic load balancing strategy.

In this paper we present a novel physical layer impairment (PLI) aware routing and wavelength assignment (RWA) approach combined with PLI-aware regenerator placement. We consider the dominant linear and nonlinear signal quality degrading effects in 10 Gb/s non-return to zero (NRZ) on-off keying (OOK) systems by analytical models for amplified spontaneous emission noise (ASE), filter crosstalk, group-velocity dispersion (GVD) and polarization mode dispersion (PMD) as well as cross-phase modulation (XPM) and four wave mixing (FWM). As a topology for our study we have chosen the COST266 reference network, which is a pan-European meshed network with a total of 28 nodes and 41 (bidirectional) edges. For this network demands have been defined based on a population-based model. To facilitate transmission over very long path lengths regenerator pools have been placed sparsely at certain nodes due to cost considerations. The regenerator sites are selected based on a heuristic algorithm taking into account the physical effects resulting in signal quality degradation. Due to the high cost of optical-electrical-optical (OEO) conversion as few regenerators as possible are deployed. In our investigations we assumed a 1+1 path protection scheme in the optical layer. The exact assessment of the signal quality based on the current traffic situation using our methods permits a network performance (wavelength blocking probability) comparable to an opaque network with only a small number of regenerators.

In this paper, we propose a novel segment-shared protection algorithm (SSPA) for multi-domain optical mesh networks. In SSPA, intra-domain links and inter-domain links are protected with segment-backup paths respectively. The segment-working path is protected with path shared protection in single-domain and the inter-domain link is protected with shared segment-backup path that it could traverse several domains. The major advantage of SSPA is its shorter recovery time than previous path protection in multi-domain optical mesh networks. The SSPA resolves the inter-domain link protection problem successfully. The computer simulation results show that resources utilization ratio of our proposed SSPA is near to the optimization algorithm that requires the global knowledge of network topology.