This invited contribution reports recent results on all-optical signal regeneration using semiconductor optical amplifiers (SOAs). Those include results obtained using the cross-gain-compression (XGC) effect in common SOAs for wavelength preserving 2R regeneration of both RZ and NRZ signals, and regenerative amplification up to 80 Gb/s obtained thanks to detuned filtering of self phase modulation (SPM) in a quantum dot (QD) SOA.

The expected development of advanced video services with HDTV quality demands the delivery of more than Gb/s link to end users across the last mile connection. Future access networks are also required to have long reach for reduction in the number of central offices (CO). Fueled by those requirements, we propose a novel equalization scheme that increases the capacity and reach of the wavelength division multiplexing passive optical network (WDM-PON) based on a low bandwidth reflective semiconductor optical amplifier (RSOA). We investigate the characteristics of 10 Gb/s upstream transmission in WDM-PON using RSOA with only 1.2 GHz electrical bandwidth and various lengths of fiber. It is proven that the proposed four-level partial response equalizer (PRE) is capable of mitigating the impact of ISI in the received signals from optical network units (ONU) located 0 km to 75 km away from the optical line terminal (OLT).

A 3-stage structure L-band EDFA is introduced to enhance the poor power conversion efficiency by utilizing the wasted ASE from the third EDF section and pump bypass in 1-2 stage. The results indicate that the EDFA could provide low-noise, high-gain and large-flatness ranging from 1570 to 1610 nm.

Wavelength conversion of RZ-OOK signals in PDM systems based on XPM in high-nonlinear-fiber (HNLF) through a polarization-diversified loop is experimentally demonstrated. Less than 1dB eye-diagram-based SNR penalty is obtained for 10-G/bs RZ-OOK PDM signals.

Radio-frequency (RF) oscillators are key components in many radio and also radio-over-fiber (RoF) systems. Here, we propose a novel technique to extent the tunable range of the radio frequency generated by optoelectronic oscillators. The proposed optoelectronic oscillator includes an intensity modulator, a photodetector, an optical circulator, a Febry-Perot laser diode, and a broad band RF amplifier. All those devices compose a single feedback loop. The key device in the system is the Febry-Perot laser. By utilizing injection locking of Febry-Perot laser diode rather than the commonly used radio-frequency band-pass filter, a high Q microwave filter is formed in optical domain. Theoretical analysis of the optoelectronic oscillator as well as the tunablity of microwave filter is provided. The central frequency of the microwave filter is a function of the optical wavelength and effective injection power, i.e. the injection power at the transverse-electric (TE) mode of the F-P laser. Thus, the central frequency of the high Q band-pass can be simply tuned by tuning the wavelength and the injection power of the injection light. Experiment is done to verify the theoretical analysis, and a tunable range from 11.074 GHz to 39.400GHz is demonstrated experimentally.

An optical OFDM system with constellation fine adjustment scheme is proposed and experimentally demonstrated. The complexity of the scheme is analyzed and proved to be acceptable. The single-mode fiber transmission distance of a 56Gb/s O-OFDM system is successfully extended by 350-km at the FEC limit with less reference symbols added.

The characteristics of commercial RSOA, i.e. fiber-to-fiber gain and small-signal electro-optical conversion (E/O) response are measured. A numerical model for RSOA including spatial variations of carrier density and ASE noise is established and validated via experimental tests. The auger recombination coefficient and active region bulk of commercial RSOA are determined by fitting the stimulated characteristics to the experimental results numerically. Performance of RSOA with different cross-sectional areas of active region is investigated based on the numerical model. The stimulated results indicate that reducing the cross-sectional area can increase its modulation bandwidth. RSOA 3-dB bandwidth is able to increase to 10.0GHz by optimizing its active region bulk.

We have already proposed the wavelength conversion method which is composed of an optical Single Side-Band (SSB) modulator and Mach-Zehnder interferometer (MZI). The optical signal is converted by using basic part of this wavelength converter with 50GHz frequency conversion. We have observed the converted frequency optical spectra, eye diagrams and bit error-rate. The converted optical signals are detected without bit error, but the received optical power at 10^-9 error rate is degraded by the conversion because of the optical amplifier noise and filtering by MZI. The wavelengths of 10Gbit/s optical packets are converted by this basic unit. The converted packets are routed using AWG de-multiplexer. The carrier components are around 15dB lower compared with signal. The switching time was 200ns which is correspond to the delay of the loop.

40-Gb/s 2-Channel all-optical 3R regeneration has been realised by a fiber optical parametric amplication(FOPA). Using degraded data signals rather than clock as the pump, all-optical 40Gb/s 2-Chanel 3R regeneration is realized with improved performance of eliminating the unequal suppression of noise between bit0 and bit1, which is different in mechanism from previous regeneration schemes. Theoretical analysis of the noise suppression mechanisms under different pump power (corresponding to "0" code or "1" code) is provided. By properly adjusting the parameters of FOPA, amplitude noise on both "0" code and "1" code can be suppressed simultaneously. The clock extraction is based on wavelenth conversion and a Fabry-Perot (F-P) filter with a high Q value of 1000. The theoretical analysis conclusions are verified through experiments, and the imbalance of the 3R regeneration using fiber optical parametric amplication is eliminated. The root mean square (RMS) jitter of the clock extracted is only 180fs. Signal to noise ratio is improved respectively from 2.7,2.4 to 4.9,5.2.

In this paper, an all-optical 1x2 switch is demonstrated based on gain modulation of injection locked Fabry-Perot laser diode (FP-LD). The control circuit generates set and reset states, which are used to perform 1x2 switch function. The switch output state is maintained once it starts to operate, whether the control signal is present or absent. In the experiment, with extinction ratio of 13.2 dB the switch output eye diagram is recorded.

We evaluate switch scale for several single-layer and hierarchical optical cross-connects. With the introduction of practical waveband add/drop ratio bound that can be derived through recently developed network design algorithms, switch scale can be significantly reduced. We identify the most cost-effective architectures by applying practical parameter values that are determined through our studies.

We propose a simple node architecture that is used to concatenate two ROADM rings. The proposed architecture employs variable filters that can effectively handle classified node input signals to realize the desired routing operations. Finally, we developed a prototype system and performed transmission experiments. The technical feasibility has been confirmed.

This paper introduces a dynamic bandwidth assignment algorithm designed for ring topology (R-DBA) in order to meet the needs of the construction of Metropolitan Area Network (MAN).This algorithm targets ring topology, based on fair iteration, supports various QoS services and has previous assignment mechanism.

A novel dynamic bandwidth allocation algorithm named GTH-DBA is proposed for metro networks. A testbed is constructed to test its performance. Numeric results show that this algorithm is effective to improve the bandwidth efficiency of the metro networks.

Successful 112 GBit/s PDM-CSRZ-DQPSK transmission with coherent detection and digital electronic DSP post processing is reported over a 1730 km SSMF plus 300 km DCF field installed DWDM-system. A modular network approach with co-propagating 10 GBit/s WDM-channels is achieved.

A simple analytical method to estimate the OSNR sensitivity of multi-level amplitude, phase and combined modulation formats is shown. The results are compared to numerical simulations with coherent and direct detection technique using RZ and NRZ pulse shape.

We show how to estimate accurately the Jones matrix of the transmission line by analyzing the Stokes vectors of DP-QPSK signals. This method can be used to perform in-situ PMD measurement in dual-polarization QPSK systems, and in addition to the constant modulus algorithm (CMA) to mitigate polarization-induced impairments. The applicability of this method to other modulation formats is discussed.

Higher-order optical multilevel signaling is a key technique for future ultra-high speed signal transmission at 100 Gbit/s or beyond. In this paper, we review various multilevel transmitter and receiver designs realizing higher-order multilevel signaling with precise optical field modulation and demodulation capabilities.

The proposed clock recovery scheme introduces electrooptical modulation to down convert the clock frequency facilitating succeeding narrow band filtering by a phase locked loop (PLL) with ordinary radio frequency (RF) devices, further, employs a quadrature phase detector in the PLL to provide an indication signal for monitoring residual dispersion. It was demonstrated in a polarization multiplexed 160-Gbit/s optical non-return to zero quadrature phase shift keying (NRZ-QPSK) transmission system.

We use a 65nm CMOS process technology to develop 40Gbit/s interface conversion prototype circuits for 40GbE, STM- 256/OC-768 and OTU3 tri-rate serial signal transport. For the first time, interface conversion functions from SFI-5.1 to SFI-5.2/XLAUI are demonstrated on a 16:4 MUX prototype chip, and from SFI-5.2/XLAUI to SFI-5.1 on a 4:16 DEMUX prototype chip. The 16:4 MUX and 4:16 DEMUX prototype chips show excellent jitter performance and consume 1.6 and 1.7 W, respectively.

We experimentally demonstrate the security vulnerability in the temporal phase coding single-user differential phase-shift keying (DPSK) and code-shift keying (CSK) OCDMA systems with a DPSK demodulator. In the experiment, we build up the 2.5Gbit/s DPSK- and CSK-OCDMA systems. In the systems, we use two 31-chip 640 Gchip/s superstructured fiber Bragg grating encoders for the signal encoding. In the receiving side, we remove the decoders and utilize the DPSK demodulator to detect the encoded signals directly. We successfully achieve the error-free BER performance and obtaine the clear open eye diagrams using the detection without the proper decoding. It indicates the existence of the eavesdropping vulnerability in the both systems. Furthermore, we also discuss the principle of DPSK demodulation attack.

This paper experimentally demonstrates the feasibility of a 2 x 2 MIMO optical communication system based on mode group diversion multiplexing in the presence of strong mode crosstalk over 250m GI-MMF. The parallel signal detection and processing are realized through the use of PD array.

We propose and experimentally demonstrate a reconfigurable all-optical VPN scheme enabling intercommunications among different ONUs in a WDM PON. Reconfiguration is realized by dynamically setting wavelength conversion of optical VPN signal using a SOA in the OLT.

A novel design of 2-to-4 photonic digital decoder has been proposed and experimentally demonstrated successfully at 40 Gbit/s. The scheme is implemented by employing differential phase shift keying (DPSK) signals, bit-inverse properties of delay interferometers (DIs), as well as cross gain modulation (XGM) of semiconductor optical amplifiers (SOAs). Correct logic results with high extinction ratios, clear and wide open eyes have been obtained simultaneously, without using any additional input beams. Furthermore, the great potential for hybrid or monolithic integration of delay interferometers and semiconductor optical amplifiers may make the scheme promising candidates for the future ultrafast photonic signal processing applications.

A compensation scheme for transmitter (Tx) and receiver (Rx) I/Q imbalance with the interference of carrier frequency offset for CO-OFDM transmission is demonstrated. It has low over-head using only two specific designed symbols for Rx side I/Q imbalance and two for that of Tx side, respectively. Rx compensation is performed in time domain after OFDM frame synchronization and the Tx I/Q imbalance is compensated in frequency domain after channel and phase estimation. The bit error rate (BER) and Q improvement are evaluated experimentally to demonstrate the effectiveness of compensation algorithms. This scheme performs better with the increase of OSNR.

We propose and experimentally demonstrate a reconfigurable two-dimensional (temporal-spectral) time domain spectral phase encoding (SPE) scheme for coherent optical code-division-multiple-access (OCDMA) application. The time-domain SPE scheme is robust to wavelength drift of the light source and is very flexible and compatible with the fiber optical system. In the proposed scheme, the ultra-short optical pulse is stretched by dispersive device and the SPE is done in time domain using high speed phase modulator. A Fiber Bragg Gratings array is used for generating the two-dimensional wavelength hopping pattern while the high speed phase modulator is used for generating the spectral phase pattern. The proposed scheme can enable simultaneous generation of the time domain spectral phase encoding and DPSK data modulation using only a single phase modulator. In the experiment, the two-dimensional SPE codes have been generated and modulated with 2.5-Gb/s DPSK data using a single phase modulator. Transmission of the 2.5-Gb/s DPSK data over 49km fiber with BER<10^-9 has been demonstrated successfully. The proposed scheme exhibits the potential to simplify the architecture and improve the security of the OCDMA system.

We investigate the transmission performance, with and without digital backpropagation, in wavelength-division-multiplexed- (WDM) transmission systems using polarisation-division-multiplexed (PDM) differentially-encoded quadrature phase shift keying (QPSK) and 16-quadrature amplitude modulation (16QAM). We consider transmission at 56GBd, 28GBd, 14GBd and 7GBd per WDM-channel while varying the channel spacing from 100GHz to 12.5GHz to maintain constant spectral efficiency per modulation format. The symmetrical split-step method based on the Manakov equation was employed to backpropagate the central WDM-channel with the help of receiver based digital signal processing. Performance of digital backpropagation was found to decrease with reduced symbol-rate, owing to the smaller proportion of the overall spectrum which is backpropagated in these cases. At higher symbol-rates digital backpropagation is up to 0.5dB more effective for 16QAM than for the QPSK format, which we attribute to the increased influence of intra-channel nonlinearities for formats with multiple intensity values leading in turn to a higher benefit when intra-channel effects are compensated for. Improvement in launch power @ BER=3x10^-3 amounts to 1.9dB for 56GBd QPSK and 2.4dB of 56GBd 16QAM.

A novel photonic method which employs the optical carrier suppression (OCS) scheme and an optical I/Q modulator to generate millimetre-wave band vector signals for 60GHz RoF systems is proposed. The generation of 2.5G Symbol/sec 16-QAM signal at 60GHz band is simulatively demonstrated as an example.

We propose and experimentally demonstrate a WDM-PON architecture supporting point-to-point, broadcast, and multicast services based on hybrid modulation format. Wireless access for multicast data is provided using optical carrier suppression (OCS) technology.

We propose a novel technique to achieve optical frequency up-conversion in millimeter wave band without any local microwave oscillator. It is hard to directly get mm band carrier frequency because the efficiency of the mode-locking decreases as the carrier frequency increases. Thanks to the dynamic charge carrier density response during the injection locking of Febry-Perot Laser diode (FP-LD), the wavelength of the leading part of carrier will be red-shifted while that of falling part will be blue-shifted. With a suitable negative dispersion device, the carrier will be pulse compressed and the 2nd harmonic of the carrier will be greatly enhanced. Thus it makes the RoF frequency up-conversion in mm band feasible. By utilizing injection locking of Febry-Perot Laser diode, a suitable negative dispersion device and polarization interferometer, a 2.5Gb/s baseband signal is frequency up-conversed to a subcarrier modulation signal at 30.8GHz.

In this paper, we propose a single sideband RoF system using high spectrum efficiency MSK modulation. In the system, one sideband of terahertz wave (THz-wave) undergoes the single sideband MSK modulation while the terahertz wave is obtained via carrier-suppression double sideband modulation. The BER performance achieves 1x10-9 and the sensitivity of the receiver is -17.8dBm with a data rate of 5Gb/s after 40km transmission.

We propose and experimentally demonstrate a simple and cost-effective bi-directional RoF system for simultaneous transmission of downstream baseband, MW and MMW signals based on a single-drive Mach-Zehnder modulator (MZM). Upstream data transmission is realized by re-modulation of downstream signals.

We propose and simulate a single-side band (SSB) Radio-over-Fiber ( RoF ) system based on OFDM-MSK modulation. In-phase/quadrature-phase (IQ) imbalance effect can significantly degrade the system performance. In order to obtain a better system performance, at the transmitter, the phase departure tolerance should be within 1 degree and the amplitude departure tolerance should be within 5% in the electrical domain, the phase departure tolerance should be within 10 degrees and the amplitude departure tolerance should be within 40% in the optical domain, the departure tolerance of splitting ratio should be within ±12%, and at the same time, time misalignment should be within ±32ps. At the receiver, the phase departure tolerance should be within 2 degrees and the amplitude departure tolerance should be within 10%. Furthermore, it's further found that the electrical part of the transmitter presents a stricter requirement on the IQ balance than the receiver, while the receiver presents a stricter requirement on the IQ balance than the optical part of the transmitter.

In this paper a novel high bandwidth optical wireless communication system for indoor personal area networking applications with both ubiquitous coverage and mobility feature is proposed and tradeoffs are studied. Gigabit-per-second optical wireless communication system with limited mobility provided to users has been demonstrated by proof-of- concept experiments for the first time. It has also been shown that when incorporating with the WiFi-based localization system, our proposed system can provide high-speed error-free operation with mobility over the entire room.

A novel data remodulation scheme with downstream Amplitude Shift Keying (ASK)- Differential Phase Shift Keying (DPSK) signals and on-off keying (OOK) upstream signal using reflective semiconductor optical amplifier (RSOA) for wavelength-division multiplexing (WDM)- passive optical networks (PON) is proposed in the letter. An experiment on 10-Gb/s DPSK with 1.25-Gb/s ASK downstream signals generation and 1.25-Gb/s upstream signal remodulation is demonstrated, and experimental results show error-free operation after ~25-km transmission can be achieved without dispersion compensation.

In incoherent-injected WDM-PONs, it is shown through simulations and experiments that Flat-band athermal AWGs with channel spacing of 100GHz have improved performance in terms of BER than their Gaussian-shape counterpart, due to the effect of signal filtering. At a BER of 10-9, the power penalty is 2 dB for Flat-band AWGs and more than 4 dB for Gaussian AWGs, which do not reach the required BER. In addition, the effect of AWG detuning in the WDM-PON systems is investigated. It is shown that by detuning the two Flat-band AWGs by 40GHz from their center wavelength, an additional 0.5dB power penalty is induced.

We propose and demonstrate a simple chirp-free optical Manchester signal transmitter that consists of a single-drive Mach-Zehnder modulator and a passive electronic power combiner. A 5-Gb/s optical Manchester signal is generated and its dispersion tolerance enhancement is investigated. The bandwidth of the modulator and driving circuit is reduced by half.

We report an experiment of de-multiplexing automatically in a polarization division multiplexing system using a DSP based de-multiplexer. The searching algorithm was a modified Particle Swarm Optimization. We successfully demonstrated the de-multiplexing scheme both in a 2x10-Gb/s PolMux-NRZ-OOK system and a 40-Gb/s PolMux-RZ-DQPSK system.

We proposed an adaptive PMD compensation scheme based on FPGA using DPSO algorithm. Stable polarization compensation for 43Gbit/s RZ-DQPSK transmission over 1200km was demonstrated with endless polarization scrambling. Excellent performance was accomplished utilizing our scheme in case of changing SOP and DGD in longhaul fibre link.

The paper presents a training sequence and an algorithm used for polarization demultiplexing in polarization division multiplexing coherent optical transmission system. The method can perform polarization demultiplex correctly and receiver is not required to synchronize with training sequence. Frequency offset can also be estimated with the training sequence.

Polarization mode dispersion is considered to be one of the main obstacles for high speed long-haul optical fiber communication systems. It is necessary to realize PMD monitoring. This paper theoretically analyzed the principle of polarization mode dispersion monitoring with degree of polarization as monitoring signal. Using the degree of polarization method, the scheme of DSP based high precision real-time inline PMD monitoring is designed and implemented. An experiment system is set up to monitor polarization mode dispersion of 40Gbit/s DQPSK system. The experiment results show that the range of input optical power is -20~0dBm, the response time is 1μs and the states of polarization accuracy degree is 1%. It has high sensitivity and can be used in a variety of modulation formats and it is independent of optical signal rate.

An optimized design of E/O and O/E converters for access points of WiMAX RoF was carried out by measuring RCE after 30 to 40km SMF. The use of the DFB, a low input-impedance driving, a low distortion PIN-PD, and a high gain EP-HEMT amplifier is promising the cost-effective design.

Photonic millimeter wave (mm-wave) signal generation employing a differential Mach-Zehnder modulator (DMZM) and a semiconductor optical amplifier (SOA) is proposed and evaluated by simulations. Optimization design and performance analysis of the scheme are conducted via simulations. This paper presents the theoretical study of the generation of copolarized dual-pump with locked phase and optical mm-wave signal, the propagation of the optical mm-wave signal in single mode fiber (SMF). The effect of SOA and the input optical signal on the mm-wave generation and transmission is investigated by means simulations. Simulation results present useful insights for the practicality of photonic mm-wave signal generation and distribution in the next-generation access network.

A novel and flexible bandwidth scheduling scheme based on three dimensional divisions multiplexing of MSK-OFDM for both downstream and upstream data in WDM-PON is proposed and experimentally demonstrated with good performances at 2.5Gbit/s by 5 sidebands.

OFDMA-PON is an attractive candidate for the next-generation PON. We study the influence of wavelength interval between adjacent ONUs in OFDMA-PON by simulation and theoretical analysis, and find the minimum wavelength intervals which are different in ideal and actual system. The ideal minimum wavelength interval is twice of OFDM symbol bandwidth, while actual minimum interval is larger than that.

Single-carrier Frequency Division Multiple Address (SC-FDMA) technique has been proposed for next generation passive optical network (PON). In this paper, we demonstrate both downstream and upstream SC-FDMA transmissions with source-free single wavelength transmission.

For the first time, we propose and demonstrate a novel three-level Signal allocation scheme OFDM PON architecture which makes the best use of OFDM characteristics. In the architecture, we produce five-carrier signal optically by MZM, use odd carriers modulating downlink data and retain even carriers for uplink data. Multiple signal distribution methods can be switched flexibly. By simulation, a small PON systems with four ONU based on this architecture was realized. Both the downlink and uplink transmission adopt 4-QAM-OFDM signal. The transmission rate will be demonstrated to exceed 10Gbit/s.

Multi- and single-carrier (SC) coherent optical fiber communication with frequency domain equalization (FDE) is discussed with emphasis on 100-Gb/s operation. 120-Gb/s coherent optical (CO-SCFDE) system over 317-km standard single-mode fiber is demonstrated. Coherent optical orthogonal frequency-division-multiplexing (CO-OFDM) and single-carrier frequency-division-multiplexing scheme (CO-SCFDM) are theoretically and experimentally compared.

Polarization-multiplexed LDPC-coded QAM robust to I-Q imbalance and polarization offset is proposed. Efficient mitigation of I-Q imbalance and polarization offset is demonstrated with LDPC-coded turbo equalization by simultaneous MAP detection of symbols transmitted over two orthogonal polarizations. The proposed scheme is much more efficient in I-Q imbalance and polarization offset compensation than conventional approaches.

Two reduced-complexity (RC) LDPC decoders are proposed, which can be used in combination with large-girth LDPC codes to enable beyond 400 Gb/s serial optical transmission. We show that optimally attenuated RC min-sum sum algorithm performs only 0.45 dB worse than conventional sum-product algorithm, while having lower storage memory requirements and much lower latency. We further evaluate the proposed algorithms for use in beyond 400 Gb/s serial optical transmission in combination with PolMUX 32-IPQ-based signal constellation and show that low BERs can be achieved for medium optical SNRs, while achieving the net coding gain above 11.4 dB.

We propose algorithm to compensate frequency offset in optical coherent MSK system with little computation requirement. Simulations are performed to demonstrate the algorithm and its performance is excellent.

The channel estimation result of non-pilot subcarriers mainly depends on adjacent pilot subcarriers. A linear interpolation algorithm has been applied to the channel estimation of non-pilot subcarriers. The performance of interpolation algorithm has been simulated in a 10Gb/s direct-detection optical orthogonal frequency division multiplexing (DDO-OFDM) communication system. Constellation convergence is greatly improved.

We investigate the problem of joint frequency offset and channel estimation for an 112Gbit/s QPSK CO-OFDM system. The paper presents a novel joint frequency offset and channel estimation method based on CP and training symbols. It is able to estimate the fractional part and integral part of frequency offset at the same time. Its estimation range is more than[-2,5 GHz,2.5 GHz] .Meanwhile, we present an efficient CD estimation and compensation method for CO-OFDM based on training symbol time-domain averaging (TSTA) and systematically investigate its robustness against transmission impairments such as the nonlinearity of MZ modulator, linewidth and frequency offset of laser. Simulation results indicate that even OSNR is very low, it estimates frequency offset and channel response very well.

Midlink optical phase conjugation (OPC) based on four wave mixing (FWM) in semiconductor optical amplifier (SOA) can be used to compensate for the nonlinearity phase noise, Gordon-Mollenauer effect, in CO-OFDM long-haul systems. Simulation of 40Gb/s CO-OFDM systems without optical dispersion compensation is presented and the Q factor can be increased by >6 dB through OPC.

Fiber nonlinearity is the important determinant of applying orthogonal frequency division multiplexing (OFDM) in optical transmission systems. In this paper, the fiber nonlinearity impairments in 40Gb/s coherent optical (CO-OFDM) system are post-compensated by a new method of fiber nonlinearity post-compensation (FNPC). FNPC located before the CO-OFDM receiver includes spectrum inversion (SI) and a subsequent 80km compensation high nonlinear fiber (HNLF), in which SI is based on four wave mixing (FWM) in semiconductor optical amplifier (SOA), not necessary to be inserted into the middle of link. The fiber nonlinearity impairments in transmission link are post-compensated after SI by transmission through the HNLF with large nonlinearity coefficient. Simulating results show that the nonlinear threshold (NLT) (for Q >10 dB) can be increased by almost 2.5 dB and maximum Q factor can increase about 1.2 dB for single-channel 40Gb/s CO-OFDM system with optical dispersion compensation. For 50-GHz channel spacing WDM systems a benefit of about 1.1 dB is still possible in NLT, equating to a 0.7 dB increase in maximum Q.

As coherent detection is employed in polarization division multiplexing coherent optical orthogonal frequency division multiplexing (PDM-CO-OFDM) system, it is important to investigate and compensate the optical phase impact on system performance. We have presented the study on the electronic pilot-aided phase compensator (PC) at receiver to compensate phase drift due to laser linewidth for 100Gb/s PDM-CO-OFDM without optical dispersion compensation in this paper. Simulating results are shown that PC is very effective and the improvement of Q is >5dB for the typical optical transport systems of 800km link with linewidth equal to 100kHz.

Use of an optical single side-band (OSSB) modulation signal engenders high bandwidth efficiency because the OSSB signal occupies only half the channel bandwidth necessary for intensity modulation. Previous reports have been clarified that the binary phase shift keying OSSB (optical BPSK-SSB) signal has relative high sideband suppression ratio and tends to degrade the detected waveform as fiber input power increases. The waveform degradation is induced by self-phase modulation (SPM) in optical fiber. In order to equalize the waveform degradation, this paper proposes the novel pre-equalization scheme that makes additional phase modulation to optical BPSK-SSB signal. The depth of phase modulation is proportional to the optical power. The equalizing method was investigated in dispersion-compensated fiber transmissions with BPSK-SSB modulation. The results prove the equalization has the possibility of 15dB improvement of fiber input power compared to the no-equalization case.

An all-optical ON-OFF switch is demonstrated using a single-mode Fabry-Perot laser diode. The operation principle of the proposed switch is the gain modulation of injection locked single mode Fabry-Perot laser diode (SM FP-LD) by an external control beam. In this paper, the operation principle is explained and a 10 Gb/s data signal is switched ON and OFF with a control signal of 200 MHz. We measured the switched output with an extinction ratio of 15.4 dB. The rising and falling times of the switch were measured as 90 ps and 80 ps, respectively.

All-optical logic NOT gate is demonstrated based on cross phase modulation in highly nonlinear fiber while using a negatively pre-chirped pulse as the probe light, with the output logic NOT signals of improved extinction ratio. Assessment of 10 Gb/s Operation is carried out using pulse probe with and without pre-chirping stage. At least 3 dB extinction ratio improvement is obtained compared to the implementation using chirp free pulse probe.

Optical switch based on quantum-dot semiconductor optical amplifier (QD-SOA) is proved to be more potential in ultra-fast optical processing. And a novel optical switch has been designed to exploit the superiorities. In the QD-SOA simulation model, we also take into account the asynchrony of gain recovery and phase recovery.

This paper describes progress in gigahertz-clocked quantum key distribution systems. It details current advances in both point-to-point and network applications. We will discuss possibilities for practical quantum key distribution using single-photon sources, and discuss the experimental system performance of GHz-clocked quantum key distribution systems focusing on issues of quantum bit error rate, net bit rate and transmission distance with different detector structures, concentrating on single-photon avalanche diode detectors, but also examining superconducting nanowire-based structures. The quantum key distribution system is designed to be environmentally robust and an examination of long-term system operation will be presented.

One 160 Gb/s optical time division demultiplexer based on two cascaded electro-absorption modulators as sampling windows was demonstrated and experimentally studied in detail. The two stages were used for the 160 Gb/s into 40 Gb/s demultiplexing and 40 Gb/s into 10 Gb/s demultiplexing respectively.

We have proposed and experimentally demonstrated an ultrafast optical pulse repetition rate multiplication technique from a relatively slow optical pulse source at 1550nm based on reconfigurable time domain spectral amplitude/phase filtering operation. In the proposed technique, a pair of dispersive fibers and a high speed electro-optical modulator driven by a 40GHz pulse pattern generator that can be rapidly programmed are used to control the repetition rate. In the experiment, repetition rate multiplication from 10GHz to a high speed repetition rate of 20GHz and 40GHz has been successfully achieved by the proposed time domain spectral amplitude/phase filtering.

Optical arbitrary waveform generation (OAWG) is considered as a promising technology in microwave photonics. Line-by- line control technique for OAWG according to Fourier transform theory has been widely adopted and then Optical frequency comb (OFC) is one key point of OAWG based on this method. In the experiment, based on cascading intensity modulation, an OFC signal of nine spectral lines is generated with its excursion smaller than 2dB. And then by selecting spectral lines of interest from OFC via optical filters, 10GHz, 20GHz sinusoidal signals with low phase noise and several desirable waveforms are generated easily.

This paper proposes a dual space diversity technique using both Space Time Block Code (STBC) and mode group diversity multiplexing (MGDM) in an 10 Gbit/s 1km long optical multi-mode fiber (MMF)communication system. The simulation results demonstrate that the system based on STBC+MGDM achieves better performance than that only employing MGDM in the current system.

We demonstrated the QPSK signal generated by a single phase-modulator (PM-QPSK) is more tolerant to nonlinearity than that generated by the Mach Zehnder Interferometer (MZI-QPSK). With simulation of WDM system at 28 Gsymbol/s per polarization the performance of PM-QPSK and MZI-QPSK are compared when dispersion management is used. After 800 km's transmission in single-polarization transmission system, MZI-QPSK outperforms PM-QPSK because of its better tolerance to ISI. After 400 km's transmission in polarization multiplexing system, PM-QPSK outperforms MZI-QPSK by comparing the symbol error ratio with optimized launch power, which is because PM-QPSK has better nonlinearity tolerance in dispersion managed system.

40 Gb/s optical demultiplexing with amplitude regeneration is demonstrated based on data pump fiber-optical parametric amplification(FOPA) while a 10 GHz short pulse is used as the probe. Four 10 Gb/s output channels with at least 14 dB extinction ratio enhancement and an average 5.5 Q factor is obtained after demultiplexing.

Network virtualization can help diversify the Internet by supporting multiple virtual network (VN) architectures on a shared substrate. Making efficient use of the underlying substrate network resources requires effective algorithms for virtual network embedding (VNE) that maps each virtual network (VN) to specific nodes and links in the substrate network. In this paper, we formulate the VNE problem as a mixed integer linear program (MILP), and then propose a progressively greedy VNE algorithm (PG-VNE) which has three main improvements over previous approaches. 1) Adding constraints after we relax the MILP to obtain STRICT_LP model which leads to better coordination between the node and link mapping stages, 2) Using the greedy idea to map virtual nodes and 3) Mapping virtual nodes progressively, i.e., mapping virtual nodes one by one. Simulation results show that PG-VNE algorithm realizes close coordination between node and link mapping stages and performs well in terms of revenue, cost and VN request acceptance ratio when compared with the well known D-ViNE and R-ViNE algorithms.

We propose and demonstrate a novel microwave photonic filter based on an electro-optic phase modulator (EOPM), a tunable optical bandpass filter (TOBF) and a cascaded fiber delay interferometer (DI). The TOBF is slightly detuned relative to the optical carrier. By using the complementary optical responses of the two output ports of DI, one port of DI is detuned the same as the TOBF and the other is detuned opposite to the TOBF. Thus a microwave bandpass filter and a microwave notch filter are obtained at the two ports of DI simultaneously. In the experiment, the rejection ratios of obtained microwave notch filter and bandpass filter are 22 dB and 16 dB, respectively.

In this paper, a tunable microwave photonic filter with one complex coefficient is proposed and experimentally demonstrated. The complex coefficient is generated using a tunable optical RF phase shifter that consists of a semiconductor optical amplifier (SOA) and a tunable band-pass filter (TBPF). The phase of the RF signal is shifted by the differentiation process in semiconductor optical amplifier (SOA) where both Cross-Phase Modulation (XPM) and Cross-Gain Modulation (XGM) effects are exploited simultaneously. A two-tap photonic microwave filter with one tunable complex coefficient and a tuning range of a full free spectral range (FSR), is experimentally demonstrated.

We demonstrate multi-channel regenerative non-return-to-zero (NRZ) to return-to-zero (RZ) conversions with tunable output pulse-width and single-to-dual function, using a phase modulator and an array waveguide grating (AWG). Transmission and bit error ratio (BER) show a good performance for the converted RZ signal compared with conventional one.

In this paper, we propose a GPON-based transmission hierarchy (GTH) for metro network. Then a simulation testbed is constructed to study the GTH's performance on supporting wireless backhaul service. Simulation results show that, compared with MSTP, GTH reduces the cost of network expansion significantly for its higher transmission efficiency.

In this paper, we propose and experimentally demonstrate a simple self-restored architecture for WDM passive optical network (PON) by only adding a novel 2x2 optical switch design in each optical networking unit (ONU). By using the proposed protection architecture, the affected traffic will be restored immediately against fiber fault in the feeder and distributed fibers. Moreover, the performance of proposed self-survivable WDM-PON is also discussed and analyzed.

In this investigation, an mm-wave RoF transmission link with dual beat-mode optical source is investigated. We demonstrate a 170 GHz band millimeter-wave source using an external dual-mode-locked Fabry-Perot laser diode (FP-LD) with a 1.38 nm mode-spacing. Two distributed feedback (DFB) lasers are used to inject the corresponding modes of FP-LD to generate dual-beat-mode output. The proposed dual-mode laser is not only stable, but also has single-longitudinal- mode (SLM) output characteristics. Moreover, we also investigate the relationship between two external-injected wavelengths for the dual-wavelength output.

Optical minimum-shift keying (MSK) format has constant envelope and continuous temporal phase shift, and exhibits a promising application for its high compact spectrum. Coherent detections have been proved to be effectively for the significant improvement of receiver sensitivity. In this paper, a heterodyne detection scheme to demodulate 40Gb/s MSK signal is proposed, and simulation analysis is given to the receiver sensitivity, the laser line-width and frequency offset. Comparison is drawn between the presented heterodyne detection and homodyne detection, and investigation is made on the performance of the receiver in different schemes.

The structure of OFDM symbol has great influence on the performance of coherent optical orthogonal frequency division multiplexing (CO-OFDM) system. We present, for the first time, a typical structure of OFDM symbol with zero padding and pilots for CO-OFDM. Virtual subcarriers is used for guard band and eliminating the impact of DC and low frequency subcarriers. Pilots are intended to track the channel variation and compensate for phase noise. Simulation results indicate that CO-OFDM system with this structure is more resistant to the influence of optical carrier, MZ modulator, linewidth, frequency offset than the system without zero padding. It is also found that 32 pilots are enough in one OFDM symbol. Furthermore, the proposed structure is very suitable for frequency offset compensation. Using the structure, performance of CO-OFDM is improved largely.

The method of coherent detection utilize image rejection in the receiver to obtain better performance in demodulation of FSK downstream. We implement it in passive optical network and adopt re-modulation which modulalte ASK signal directly on FSK signal as upstream. The possibility of extending the system to WDM-PON condition is also discussed. By image rejection receiver, we get obvious performance improvement featuring BER. When in long haul situation, compared with band-pass filter, the image rejection receiver greatly improve the BER. Several detailed parameters in the receiver are varied in certain range to indicate their effect on the system performance. Results show that the image rejection receiver we propose is suitable for access network, especially for long-distance access network.

Radio-over-fiber (RoF) has been proposed as an enabling technology for broadband networks, such as WiMAX and WiFi. Besides the inherent high bandwidth and reliability of RoF systems, they also allow the reduction of installation and maintenance cost of the remote antenna units (RAUs) and improvement in the coverage area of the base station/access point. However, the nonlinear distortion of the optical link, which stems mainly from the laser diode, may impose serious limitations on the system performance, especially when high PAPR, wideband signals are used. Thus, distortion compensation is a key issue in order to facilitate the application of the RoF technology for broadband networks. In this work, digital predistortion for directly modulated RoF links is experimentally investigated. A memory-polynomial- based predistorter model is identified from measurements of the baseband OFDM input-output signals and through the use of an indirect learning architecture. Then, a predistorted signal is generated and fed to the RoF link for comparing its output with that of the non-predistorted one. As a result of this compensation technique, an improvement of the static link linearity and the output constellation diagram has been found, with an EVM reduction of 1.73 %.

Further increase of transmission capacity is required for high speed networks and there are some problems to improve WDM system performances. Especially, for-wave-mixing (FWM) is one of the fiber nonlinearities and FWM impairments remain unsolved. In this experiment, we demonstrate digital coherent detection for FWM compensation. Two ASK-modulated optical signals are transmitted through dispersion-shifted fibers (DSF) with an optical amplifier repeater. FWM has been generated by transmitting through DSF. The transmitted signals are combined with phase-locked local oscillator lights because phase information and the phase relation among individual channels can be kept using phase locked light. The detected individual signals are compensated by backward propagation method. The FWM-crosstalk ratio is improved from 12 dB to 17dB and eye-opening is also improved by compensation.

The proposed maximum likelihood sequence estimation (MLSE) receiver to mitigate polarization mode dispersion (PMD) impairment on phase-shift keying incorporates polarization diversity receiving and joint symbol processing by employing another delay interferometer (DI) using cross polarization interference together with a traditional DI. Applied to an 80-Gbit/s DQPSK demonstration system, the joint-symbol MLSE receiver shows better performance.

We have proposed a modified all-optical SPM-based 2R-regenerator using an imbalanced nonlinear optical loop mirror (I-NOLM). The benefits of using the I-NOLM are demonstrated. Simulation and experiment of 10-Gbit/s regeneration are also performed. Research results show that Q-factor and extinction-ratio of the degraded signal can be improved more with the modified regenerator compared to a conventional SPM-based 2R-regenerator.

This paper reports the prototype of optical polarization mode dispersion compensator (OPMDC) we have made to compensate the polarization mode dispersion (PMD) in fibers. The OPMDC prototype designed was tested in a 43Gb/s RZ-DQPSK system either in the back-to-back case, or in a 1200km transmission testbed. It showed a good performance under the tests of fast SOP and PSP rotation, DGD variation, and moderate knock on the testbed in a period of 12 hours.