Search

In this work, we propose a new scheme to generate high-quality frequency 16-tupling millimeter-wave only by two Mach-Zehnder modulators (MZM) and without filter. The two MZMs, biased at maximum transmission point, are used for odd order sidebands suppression. By controlling the modulation index of the MZM, ±4th-order sidebands amplitude return zero. By adjusting phase shifter and the attenuator to cancel 0th-order sideband; By adjusting the phase difference between RF driving signals of two subsystems to cancel ±(4n-2)th-order optical sidebands. The feasibility of the proposed were demonstrated and discussed. Simulation results show that using 10 GHz RF signal to drive the MZMs, 160 GHz mm-wave signal with 31.35 dB optical sideband suppression ratio (OSSR) and 24.11 dB radio frequency sideband suppression ratio (RFSSR) was obtained. The results are coincide with the theoretical analysis. Furthermore, the non-ideal cases are also analyzed. Even if the modulation index, the MZM bias voltage or extinction ratios deviate from the ideal values to a certain degree, the mm-wave signals’ performance are still acceptable.

The specially designed Kalman filter was proved to be competent to complete the polarization equalization in some extreme environments with the large time-varying polarization mode dispersion (PMD) and the ultrafast rotation of state of polarization (RSOP). But the performance of this kind of Kalman filter is strongly restricted by the initial noise statistics represented by the process noise covariance Q and the measurement noise covariance R which will limit its effectiveness in an actual scene. To solve this problem, this paper proposes an adaptive Kalman filter structure by an adjustable way of estimating the process noise covariance Q and the measurement noise covariance R based on the measurement residual in order to be adaptive to the time-varying environments. The performance of the proposed adaptive Kalman scheme is checked both by the simulation and experiment on the 28 Gbaud PDM-QPSK/16QAM coherent optical communication system platforms. The simulation and experiment results demonstrate that the proposed adaptive Kalman filter scheme has a stable and better performance of the equalization for large PMD combined with ultrafast RSOP, with a greater tolerance the to time-varying extreme polarization environments than that using the extended Kalman filter scheme adopted in our previous work.

We present a novel hole-assisted elliptical-core polarization-maintaining few-mode fiber (EC-PM-FMF) composed of three layers of air holes. The proposed fiber breaks degeneracy of modes by introducing geometry asymmetry. Through optimizing the parameters, the designed fiber could support 18 separate polarization-maintaining modes with the minimum effective refractive index difference ( Δ n e f f ) between adjacent modes larger than 1.21 × 10−4 over the whole C + L band. Moreover, the designed fiber exhibits good mode and propagation properties including large effective mode area, low nonlinear coefficient, small and flat chromatic dispersion, low confinement loss and bending loss. Thus, the fiber can be used in mode division multiplexing (MDM) system without multi-input multi-output (MIMO) digital signal processing (DSP) technology to improve channel utilization.

Based on 5 requirements which are essential for stable OAM mode transmission, we propose an OAM fiber family based on a structure of circular photonic crystal fiber (C-PCF). The proposed C-PCF in the family is made of pure silica, with a big round air hole at the center, several rings of air-hole array as the cladding, and a ring shaped silica area in between as the core where the OAM modes propagate. We also provide a design strategy with which the optimized C-PCF can be obtained with optimum number of high quality OAM modes (up to 42 OAM modes), large effective index separation for corresponding vector modes over a wide bandwidth, relative small and flat dispersion, and low nonlinear coefficient compared with a conventional single mode fiber. The designed fiber can be used in MDM communications and other OAM applications in fibers.

A joint carrier recovery scheme for polarization division multiplexing (PDM) coherent optical transmission system is proposed and demonstrated, in which the extended Kalman filter (EKF) is exploited to estimate and equalize the carrier frequency offset (CFO) and carrier phase noise (CPN) simultaneously. The proposed method is implemented and verified in the PDM-QPSK system and the PDM-16QAM system with the comparisons to conventional improved Mth-power (IMP) algorithm for CFO estimation, blind phase search (BPS) algorithm or Viterbi-Viterbi (V-V) algorithm for CPN recovery. It is demonstrated that the proposed scheme shows high CFO estimation accuracy, with absolute mean estimation error below 1.5 MHz. Meanwhile, the proposed method has the CFO tolerance of [±3 GHz] for PDM-QPSK system and [±0.9 GHz] for PDM-16QAM system. Compare with IMP/BPS and IMP/V-V, the proposed scheme can enhance the linewidth symbol duration product from 3 × 10−4 (IMP/BPS) and 2 × 10−4 (IMP/V-V) to 1 × 10−3 for PDM-QPSK, and from 1 × 10−4 (IMP/BPS) to 3 × 10−4 for PDM-16QAM, respectively, at the 1 dB optical signal-to-noise ratio (OSNR) penalty. The proposed Kalman filter also shows a fast convergence with only 100 symbols and much lower computational complexity.

We propose a promising fiber nonlinearity-tolerant optical signal-to-noise ratio (OSNR) monitoring technique for high speed long haul polarization multiplexed-quadrature phase shift keying (PM-QPSK) Nyquist wavelength division multiplexing (WDM) coherent optical fiber transmission system, in which the correlation function between the symbols in all states of polarization as well as neighboring channels are exploited instead of the traditional GN model, thus it does not need the prior knowledge of the fiber transmission link nor the complicated integrals. The method is demonstrated to have the OSNR measurement error below 1.28 dB over a wide OSNR range from 10 dB to 28 dB in 100 Gb/s per channel PM-QPSK-Nyquist-WDM system when the launch power is up to 4 dBm and transmission distance is as long as 1600 km.

A generalized Burgers equation with variable coefficients is introduced based on the (2+1)-dimensional Burgers equation. Using the test function method combined with the bilinear form, we obtain the lump solutions to the generalized Burgers equation with variable coefficients. The amplitude and velocity of the extremum point are derived to analyze the propagation of the lump wave. Moreover, we derive and study the mixed solutions including lump-one-kink and lump-two-kink cases. With symbolic computation, two cases of relations among the parameters are yielded corresponding to the solutions. Different and interesting interaction phenomena arise from assigning abundant functions to the variable coefficients. Especially, we find that the shape of kink waves might be parabolic type, and one lump wave can be decomposed into two lump waves. The test function method is applicable for the generalized Burgers equation with variable coefficients, and it will be applied to some other variable-coefficient equations in the future.

The (2+1)-dimensional Kadomtsev-Petviashvili type equations describe the nonlinear phenomena and characteristics in oceanography, fluid dynamics and shallow water. In the literature, a novel (2+1)-dimensional nonlinear model is proposed, and the localized wave interaction solutions are studied including lump-kink and lump-soliton types. Hereby, it is of further value to investigate the integrability characteristics of this model. In this paper, we firstly conduct the Painleve analysis and find it fails to pass the Painleve test due to a non-vanishing compatibility condition at the highest resonance level. Then we derive the soliton solutions and give the formula of the N -soliton solution, which is proved by means of the Hirota condition. The criterion for the linear superposition principle is also given to generate the resonant solutions. Backlund transformation, Lax pair and infinitely many conservation laws are derived through the Hirota bilinear method and Bell polynomial approach. As a result, we have a more overall understanding of the integrability characteristics of this novel (2+1)-dimensional nonlinear model.

A new method of optical signal-to-noise ratio (OSNR) monitoring in presence of fiber nonlinearities, suitable for Nyquist Wavelength Division Multiplexing (Nyquist-WDM) coherent optical transmission system, is proposed, in which the correlation function (CF) between the symbols who come from neighboring channels is used. The effectiveness of the proposed method is numerically verified in a 28 Gbaud per channel Nyquist-QPSK-WDM coherent system. The results show that, under the configuration with fiber nonlinearities, this method can achieve OSNR monitoring with the error less than 1.5 dB in the OSNR range of 10–24 dB, and keep valid regardless of the transmission distances, channel input powers and number of WDM channels.

A new design of a circular photonic crystal fiber, which can support 26 OAM modes, is proposed. The numerical analysis shows that the proposed fiber possesses very good values of the fiber parameters, such as low confinement loss which is lower than 0.003 dB/m at wavelength of 1.55 μm, flat chromatic dispersion whose dispersion variation for the OAM modes with the order number | l | ⩽ 4 is within 39.84 ps/nm/km over a 750 nm bandwidth from 1.25 to 2 μm, and small nonlinear coefficient which is less than 2.08 W−1/km for all modes at 1.55 μm. In addition, the influence of geometric deformation on the modal birefringence of the proposed fiber has also been discussed. It is proved that the modal birefringence caused by the geometric deformation is different for different vector modes, and most of the higher order modes have a smaller modal birefringence than the lower order modes when the ellipticity is less than 1.08.

In this paper, a novel and relatively simple scheme for generation of high-quality 16QAM-modulated multi-carrier vector millimeter-wave (mm-wave) was proposed, which was only based on a single Mach–Zehnder modulator (MZM), while the precoding technique and optical filter were not needed. At the transmitter, electrical dual single sideband (SSB) signals were generated by a digital signal processing (DSP) module, in which the frequency of modulated signals could be adjusted flexibly. Employing the generated dual-SSB signal to drive a push–pull MZM biased at Quad point, optical dual-SSB 16-QAM OFDM signals were obtained. After a photodiode, higher frequency OFDM vector mm-wave signals were achieved by beating. Based on the proposed method, a simulation platform of generating 30 GHz 16QAM OFDM vector mm-wave signals was set up and the performance of the generated OFDM vector mm-wave was analyzed. The results show that the bit-error ratio of the transmission system were less than hard decision forward-error correction threshold of 3 . 8 × 1 0 − 3 , which demonstrates that the generation of OFDM vector signal can work well.

A novel and relatively simple scheme for the generation of high-quality V-band (50–75 GHz) OFDM vector millimeter-wave (mm-wave) is presented, which is only enabled by a dual-parallel Mach–Zehnder modulator (DP-MZM). In this scheme, precoding technique and optical filter are not needed. At the transmitter, electrical dual single sideband (SSB) signals are generated by a digital signal processing (DSP) module, in which the frequency of modulated signals could be adjusted flexibly. The two push–pull sub-MZMs are biased at quadrature transmission points to generate dual optical single sideband (OSSB) signals. The DP-MZM is biased at null transmission point to suppress the center optical carrier and an asymmetrical OSSB signal can be obtained. Then, a higher frequency OFDM vector mm-wave signal is generated after the beating of the asymmetrical OSSB signal at the photodiode. Theoretical analysis is given out in the paper, and V-band OFDM vector mm-wave at the frequency of 65 GHz is generated in the simulation. The results show that the bit-error ratio (BER) of the transmission system is smaller than forward-error correction (FEC) threshold of 3 . 8 × 1 0 − 3 , when the optical power into photodiode (PD) of 10/20/30 Gbaud cases are not less than −19.13 dB, −17.81 dB, −16.50 dB respectively after 50km single model fiber transmission.

Inferring home locations for users from spatiotemporal data has become increasingly important for real-world applications ranging from security, recommendation, advertisement targeting, to transportation scheduling. Existing home location inference studies are based either on geo-tagged social media data or continuous GPS data. Yet this inference problem in highly sparse vehicle trajectories in urban surveillance systems remains largely unexplored. In this paper, we propose an accurate home location inference framework for vehicles in urban traffic surveillance systems by considering both spatial and temporal characteristics. To the best of our knowledge, we are the first to predict exact home community for vehicles at such a fine granularity using the sparse and noisy surveillance camera data. First, we collect and preprocess multiple contextual datasets to obtain a context-rich road network with residential communities and surveillance cameras. Second, we detect the potential home location areas for each vehicle by clustering Origin–Destination ( O - D ) pairs extracted in vehicle’s camera-based trajectories. Then we further propose an i n ∕ o u t time pattern to distinguish the home area candidate from the O - D clusters by leveraging time-aware constraints. Furthermore, to find the exact home community, we propose a Kernel Density Estimation (KDE) based inference method with a local camera selection strategy to effectively identify the home community from the residential communities near/in the home area candidate. Our comprehensive experiments on a large-scale real-world dataset demonstrate the effectiveness of our proposed method.

Unsupervised techniques typically rely on the probability density distribution of the data to detect anomalies, where objects with low probability density are considered to be abnormal. However, modeling the density distribution of high dimensional data is known to be hard, making the problem of detecting anomalies from high-dimensional data challenging. The state-of-the-art methods solve this problem by first applying dimension reduction techniques to the data and then detecting anomalies in the low dimensional space. Unfortunately, the low dimensional space does not necessarily preserve the density distribution of the original high dimensional data. This jeopardizes the effectiveness of anomaly detection. In this work, we propose a novel high dimensional anomaly detection method called LAKE. The key idea of LAKE is to unify the representation learning capacity of layer-constrained variational autoencoder with the density estimation power of kernel density estimation (KDE). Then a probability density distribution of the high dimensional data can be learned, which is able to effectively separate the anomalies out. LAKE successfully consolidates the merits of the two worlds, namely layer-constrained variational autoencoder and KDE by using a probability density-aware strategy in the training process of the autoencoder. Extensive experiments on six public benchmark datasets demonstrate that our method significantly outperforms the state-of-the-art methods in detecting anomalies and achieves up to 37% improvement in F 1 score.

A polarization demultiplexing (PolDemux) scheme using extended Kalman filter (EKF) is proposed to work in probabilistic shaping Stokes vector direct detection (PS SV-DD) system which has a good performance in the rotation of state of polarization (RSOP) compensation with lower optical signal to noise ratio (OSNR) requirement. As the ideal constellation points for the probabilistically shaped (PS) signal quite distinguished from the uniformly distributed signal in Stokes space, the algorithm of PolDemux will be somewhat different. To handle this problem, we propose a PolDemux scheme which can be suitable for both uniformly distributed quadrature amplitude modulation (QAM) signal and PS QAM signal. A 28GBaud Nyquist PS SV-DD system is constructed for performance evaluations. Different PolDemux algorithms are implemented for comparisons. Four different PS 16-QAM signals with different probability distributions and a PS 64-QAM signal with the entropy of 4bit/symbol are used to check the performances in the PS SV-DD system. The simulation results verify that the proposed scheme has the capability of RSOP tracking speed up to 41Mrad/s. Moreover, the proposed scheme performs better comparing with other algorithms, no matter what RSOP impairment model is chosen, either the model with 2 parameters or a universal model with 3 parameters compared with other PolDemux algorithms.

We propose and analyze a design of multi-OAM-modes ring-core fiber with two guided modes regions which possess relatively large effective index separations required for the vector modes transmission. The proposed fiber can support up to 28 information states bearing OAM spanning 8 OAM orders in the ring region, and two degenerate fundamental polarization modes in the core region across the whole C bands. Fiber features such as dispersion, differential mode delay, effective mode area, power isolation between two guided regions and modal birefringence have been analyzed in this paper. For the reason that the proposed fiber possess the relatively good features, it has potential applications in the next generation fiber communication systems either in the quantum domain or in the classical domain.

A watermarking LDPC code, which is a strategy designed to improve the performance of the traditional LDPC code, was introduced. By inserting some pre-defined watermarking bits into original LDPC code, we can obtain a more correct estimation about the noise level in the fiber channel. Then we use them to modify the probability distribution function (PDF) used in the initial process of belief propagation (BP) decoding algorithm. This algorithm was tested in a 128 Gb/s PDM-DQPSK optical communication system and results showed that the watermarking LDPC code had a better tolerances to polarization mode dispersion (PMD) and nonlinearity than that of traditional LDPC code. Also, by losing about 2.4% of redundancy for watermarking bits, the decoding efficiency of the watermarking LDPC code is about twice of the traditional one.