KEYWORDS: Interference (communication), Signal filtering, Covariance matrices, Control systems, Mathematical optimization, Detection and tracking algorithms, Statistical analysis, Signal detection, Matrices, Signal processing
The Stochastic Parallel Gradient Descent (SPGD) algorithm is a widely used model-free optimization algorithm in adaptive optical systems. However, when affected by noise, the SPGD algorithm may incorrectly estimate the gradient, resulting in a significant decrease in its performance and limitations in its application in complex environments. To address this issue, this paper introduces the optimal estimation algorithm and adds a self-checking process to the SPGD algorithm, proposing a Kal-SPGD algorithm. The feasibility of the Kal-SPGD algorithm is verified through simulation and experimentation. The experimental results show that compared with the standard SPGD, the use of the Kal-SPGD algorithm can reduce the signal amplitude variance by up to 92.15%. The Kal-SPGD algorithm provides a more favorable control technique for adaptive optical systems.
Low-time jitter optical pulses are widely used in precision ranging, photonic microwave generation, optical frequency comb generation, and photonic sampling. In order to obtain highly stable optical pulses at 1560 nm with an ultra-high pulse frequency of approximately 10GHz, a self-regeneratively ultrafast mode-locked laser with a polarization-maintaining fiber cavity is demonstrated in this paper. The laser adopts a section of polarization-maintaining erbium-doped fiber as the gain medium, a lithium niobate phase modulator as the active modulation device, and the rest of the resonant cavity are composed of passive polarization-maintaining fibers. This results in a total ring cavity length of about 10.7 m. The ring cavity can stimulate multiple longitudinal modes under free running conditions. One of these longitudinal mode frequencies is selected through using a self-built clock extraction and recovery module to generate RF signal with a frequency of 10 GHz. The amplified RF signal at 10 GHz drives the phase modulator and modulates the optical field in the cavity, which results in a stable self-regeneratively mode-locked pulse in the entire laser loop. It quickly achieves a stable ultrafast mode-locked state with a low timing jitter without any external RF reference. A pulse frequency of 10.0076 GHz and a pulse width of 3.14 ps were obtained, together with a side-mode suppression ratio of more than 80 dB and a phase noise of about -110 dBc/Hz@10 kHz. The characteristics of this laser, such as long-time stability, repetition rate, and spectral stability, are investigated in detail. Besides, some typical lasing states in experiments are compared and analyzed.
Chip-scale narrow-linewidth lasers have rich applications in sensing, communication, spectroscopy and light detection and ranging (LiDAR). Self-injection locking is one of the most efficient techniques to reduce linewidth significantly. By locking a laser to an external cavity, some amounts of light reflect back into the laser for mode competition, leading to a significant reduction of the lasing linewidth. In this work, we demonstrated a hybrid-integrated laser with a Microring Resonator (MRR) butt-coupled to a Distributed Feedback (DFB) laser. The radius of the MMR is designed to be 442.3 μm, corresponding to a Free Spectral Range (FSR) of about 50 GHz. And the MMR has a quality factor (Q factor) of 3×106 , fabricated in an ultralow loss silicon nitride (Si3N4) waveguide platform. In this way, the frequency noise has been reduced to 12.565 Hz²/Hz at the 10 MHz offset frequency. Finally, 40 Hz intrinsic linewidth and 91.2 kHz integral linewidth are achieved, characterized by a delayed self-heterodyne interferometer.
The spectral range of the optical frequency comb (OFC) generated by electro-optic modulators (EOMs) is limited by the modulation depth. In this paper, an on-chip system based on a Si3N4 waveguide is built to broaden a 12.5 GHz electrooptic comb (EO-comb). A numerical simulation is carried out to demonstrate the nonlinearities of self-phase modulation (SPM), Raman soliton self-frequency shift, and dispersive wave generation dominating the spectral broadening of the EO-comb in the waveguide.
In order to overcome the bottleneck of digital radio-frequency (RF) memory such as narrow instantaneous bandwidth and low analogue-digital conversion rate, photonic RF memory technology with broad instantaneous bandwidth and fast response capability has been developed. In this paper, RF memory structure based on active recirculating loop with depth and high stable signal-to-noise ratio(SNR) is proposed and experimentally verified, which can meet the requirements of bandwidth and response speed in current RF storage field. This structure combines the active optical switch with the fiber delay-loop-based structure, which greatly increases the storage duration, suppresses out-of-pulse noise and improves the SNR while ensuring the ultralow loss of the photonic loop. The structure realizes RF storage with real-time response in 2- 18GHz frequency spectrum and pulse replication times is more than 360(equal to 18μs delay time), the average level of SNR is 20dB and its fluctuation over time is less than 1.5dB, which can be applied to different applications where broad transient bandwidth storage with long time delay is needed.
We demonstrated the application of optical frequency combs in the realization of programmable microwave photonic filters. A reconfigurable high-order PF filter based on a wide-bandwidth optical frequency comb was designed, and the increasing of the tap (comb tooth) numbers results in the increasing of the order of the microwave photonic filter. In this article, the repetition frequency of the optical frequency comb is 12.5 GHz, and the optical spectral coverage reaches 25.6 nm. The order of the RF filter obtained in the experiment reaches 256, the free spectrum range (FSR) reaches 11 GHz, the 3dB-passband width is from 300 MHz to 2.6 GHz, and the sidelobe suppression is as high as 26 dB. A progressive optical pulse shaper was used to program the tap weights, which allow us to shape the bandpass of the filter. In addition, by adjusting the spectral distribution of the optical frequency comb, two types of RF filters with different passband shapes can be realized, namely, Gaussian and Sinc filters.
CMOS based Kerr soliton frequency comb has been demonstrated as the most promising multi-wavelength onchip light source. In this work, basic parameters of a Si3N4 micro-ring resonator (MRR) are characterized via the assistance of a fiber ring resonator and the calibration of a gas cell. Besides, by using of an auxiliary laser to suppress thermal dragging dynamics in dissipative soliton comb formation, a stable 100GHz Kerr soliton frequency comb generation is achieved.
An ultrafast high-repetition-rate mode-locked fiber laser based on a semiconductor saturable absorption mirror is reported in this paper. The repetition rate is as high as 1 GHz. A piece of single-mode EDF is used as the gain medium and a linear cavity is built. The central wavelength of the mode-locked fiber laser is 1561 nm, the spectral width is 10.10 nm, the pulse interval is 1ns, the RF signal-to-noise ratio is greater than 68 dB, and the pulse width is 247 fs. The experimental results show that the stable high repetition rate mode-locked pulse can be obtained through this simple and compact structure, and the application of the mode-locked laser in optical sub-sampling is verified.
In this paper, a broadband microwave photonic channelized receiver based on optical frequency comb (OFC) injection locking technology is illustrated. The simulation results show that, with more than 80 comb lines generated from OFC, this receiver enables channelized scanning and reception of broadband signal up to 40 GHz with instantaneous bandwidth of 1 GHz. Meanwhile, the channels selected using optical injection locking (OIL) technology, perform high gain and low phase noise with suppression ratio between the selected comb line and other comb lines is 28.7 dB. Due to OIL technology, the wideband tunability of this receiver would not depend on the optical filter or demultiplexer, and the band limitation and operation resolution introduced by optical components are broken through. The OIL technology also lead to the architecture of receiver more compact and feasible in practical.
In this paper, we illustrate a discrete Fourier processor based on dual optical frequency combs (OFCs). The simulation verified the DFT calculation with the frequency coverage of 0.2 GHz to 3 GHz and the resolution is 200 MHz. Meanwhile, demodulation of standard OFDM signal in time domain is simulated. This scheme avoids the problem of "electronic bottleneck" in current all-electronic processors. It is superior to the existing photon-assisted Fourier transform schemes in flexibility and accuracy, and has the characteristics of low power consumption, high throughput and high precision.
The spatial resolution of hyperspectral data is low, so there are a large number of mixed pixels, which is also one of the main reasons that reduce the accuracy of hyperspectral image target classification. Hyperspectral unmixing is an important subject in the field of remote sensing. Hyperspectral unmixing generally consists of three steps: reduction, endmember extraction and inversion. As one of the key steps of hyperspectral unmixing, efficient and rapid endmember extraction is an important object in hyperspectral remote sensing. In this paper, the endmember extraction of hyperspectral data is implemented based on PCA and a new SGA algorithm, which solves the dimension limitation of traditional SGA algorithm and the new SGA algorithm without data redundancy caused by data dimensionality reduction.
KEYWORDS: Telecommunications, Data processing, Image processing, Human-machine interfaces, Algorithm development, Hyperspectral imaging, Data acquisition, Data conversion, Data communications, Imaging systems
With the continuous expansion of hyperspectral application scenarios, the traditional universal hyperspectral data processing software system is difficult to meet the needs of the industry, and cannot be quickly connected to the intelligent processing algorithm developed by the industry, which has become one of the bottlenecks in the promotion of hyperspectral to practical applications. In order to meet the needs of various industries for professional processing of hyperspectral data, fast access to intelligent processing algorithm, and highly efficient and reliable transplantation of intelligent processing algorithm to airborne platform, this paper designs an airborne hyperspectral data processing platform compatible with intelligent processing algorithms. The software architecture of "Platform + Plug-in" is realized, which provides comprehensive support for hyperspectral image processing and enables users to focus on the development of intelligent processing algorithms, which can be compatible with different intelligent processing algorithms through simple configuration.
Deep residual networks (ResNets) can learn deep feature representation from hyperspectral images (HSIs), and therefore have been widely used for HSI classification. Despite their high accuracies, there still exist a lot of challenging cases, such as open world recognition, limited-sample learning and visualization of learned classification features, which cannot be well addressed. Most of the challenges in HSI classification can be attributed to the dependence on softmax based loss function and classifiers, which cause the lack of robustness for deep learning models and the hardness to visualize the learned classification features. To improve the robustness and achieve the visualization of learned classification features, we propose a novel learning framework called Residual Prototype Learning Network, a combination of residual network and prototypes learning mechanism. Under the framework, a prototype learning based loss function is proposed to enhance intra-class compactness and the inter-class separation of these feature representations; in addition, a prototype learning based classifier is simultaneously proposed to achieve the 2D or 3D visualization of the classification features. The effectiveness of our proposed learning framework is evaluated on several publicly available HSI benchmarks, and the experimental results show that our approach achieve better results than traditional softmax based ResNets.
Researches have shown that using convolution neural network (CNN) on spatial-spectral domain can improve the performance of hyperspectral image (HSI) classification in recently years. However, due to the existence of spectral redundancy and the high dimensional kernels used in 3D-CNN, the HSI classification models are often heavy with a huge number of parameters and high computation complexity. Motivated by the lightweight model, this paper introduced a modular convolution structure named three-dimensional interleaved group convolution (3D-IGC). This structure contains two successive group convolutions with a channel shuffle operation between them. First group convolution extracts feature on spatial-spectral domain. Then the channel shuffle enables cross-group information interchange. After this, the second group convolution perform the point-wise convolution. We proved that an IGC is wider than a normal convolution in most cases by inferred formula. The empirical results demonstrate that the increment of width in 3D-IGC model is beneficial to HSI classification with the computation complexity preserved, especially when the model has fewer parameters. Compared with the normal convolution, the 3D-IGC can largely reduce the redundancy of convolution filters in channel domain, which greatly decreases the number of parameters and the computation cost without losing classification accuracy. We also considered the effects of the 3D-IGC on deep neural networks, therefore we used the 3D-IGC to modify the residual unit and get a lightweight model compared with ResNets.
Traditional hyperspectral feature extraction methods focus on spectral features and neglect spatial features,its extraction method is set in advance and is not suitable for all hyperspectral images. Faced with these problems, we propose a three-dimensional convolutional network for hyperspectral classification, which consists of a convolutional layer,2 downsampling layers, 2 identification layers, a flatten layer, and 4 fully connected layers. The proposed network employs three-dimensional convolution operation to extract spectral-spatial features from hyperspectral images,there are two reasons for this, the first reason is three-dimensional convolution can automatically learn a large number of mappings between input and output.The second reason is three-dimensional convolution can effectively extract spectral-spatial features and improve network classification performance. In order to extract high-level features and prevent network performance degradation, the proposed network adopts residual connections.More importantly, the OpenMax algorithm is employed to detect hyperspectral unknown targets. In addition to the probability that the output belongs to a known class, the OpenmMx adds the probability that the predicted input belongs to unknown classes, as a result,the deep convolutional network can respond to inputs of unknown classes.experiments based on typical hyperspectral data show that the proposed network perform accurately in the known classes classification and the openmax algorithm is suitable for unknown targets detection of hyperspectral images.
In this paper, we illustrate a radio-frequency (RF) channelization scheme based on dual optical frequency combs (OFCs). Modulated by two sets of cascaded electro-optic (EO) modulators, dual coherent OFCs with different free spectrum ranges (FSRs), namely 10 GHz and 11 GHz, are generated. Based on OFCs, a channelizer with six channels, 1 GHz channel spacing is experimentally demonstrated, covering frequency range from 1 GHz to 7 GHz. The input RF signal is impressed onto the signal comb by a Dual-Parallel Mach-Zander modulator (DPMZM). At the same time, balance detection methods and I/Q demodulation are used to achieve high-precision reception and the processing of RF signals. Selecting the corresponding channel with a waveshaper, the performance of six channels is shown and the spectra at different stages are also demonstrated.
Signal to noise ratio (SNR) is one of the key parameters in the communication, radar and spectrum perception systems. In this paper, we propose and demonstrate a SNR enhancement receiver with wide processing bandwidth and tunability, where two coherent optical frequency combs (OFCs) based on multi-channel microwave source and electro-optic modulators are incorporated to accomplish simultaneous frequency down-conversion and channelization. By exploiting a dual frequency microwave source as a comb driver, the FSR tunable OFC is demonstrated. The FSR of the laser can be tuned flexibly from 8 GHz to 12 GHz by controlling the controlling the frequency of the microwave source. Multifrequency microwave signal is generated from a 0-10GHz microwave signal source and cloned to the optical domain by a carrier suppressed single sideband modulator (CS-SS) and then down-converted and channelized to the same IF. The IF signals are digitalized and then added in the digital domain. With the proposed receiver a 7.2 dB SNR enhancement has been achieved.
The exciton complexes in two-dimensional materials have long fascinated scientists and researchers for their mechanisms in fundamental photo-physics. And it is well established that the evolution of defect bound excitons in twodimensional semiconducting TMDs brings largely unexplored opportunities for tailoring their optoelectronic properties. Yet thus far, the properties of defect bound excitons of TMDs have been rarely investigated. In this work, the intrinsic properties of defect bound excitons in aged CVD-grown monolayer WS2 are experimentally studied by the steady-state photoluminescence measurement. Specifically, the photoluminescence mapping experiment is conducted to demonstrate the spatial distribution of the defect bound excitons, whose spectral feature is located ~0.2 eV below the neutral free Aexcitons. Additionally, the power-dependent photoluminescence experiment is applied to investigate the behavior of the defect-state photoluminescence and a significant nonlinear dependence of defect bound excitons on excitation power is revealed. Furthermore, we directly observed the disappearance of defect-state photoluminescence by exposing sample to high laser power irradiation, which can be explained by the enhanced desorption process of molecules physiosorbed on surfaces under laser irradiation. The results of our work provide a comprehensive understanding for the defect bound excitons in monolayer tungsten disulfide, which is essential in promoting the development of defect engineering about two-dimensional semiconducting TMDs and may pave the way for tailoring the performance of the optoelectronic device.
In this work, few layers of Bi2Se3 is chemically treated, in which the AuCl3 solution is used for oxidation reaction to form p-doping, and BV solution (benzyl dichloride) is put to form n-doping to change material properties. We used pumpprobe system to verify the effect of doping on Bi2Se3 materials. In addition, the nonlinear saturable absorption of the material is also controlled. Through the I-scan test, we found that the saturable absorption has diverse responses to different wavelengths and doping conditions. By doping, the Fermi level of the material can be adjusted to control the saturable absorption of the material, which can be applied to the mode-locked laser. The weakened saturable light intensity can make the mode-locked pulse easier to generate.
The optical filters are studied according to the applications of photonic-assisted radio frequency (RF) signal gereration and all optical coherent signal processing. Firstly, the finite impulse response (FIR) filter is analyzed theoretically. And it points out that the advantages of FIR lies in the reconfiguration, real coefficients and linear-phase impulse response which could guarantee the coherence of the ultrashort pulse. Based on this, the specific filters for free spectral range (FSR) broadening and sinusoidal signal generation are designed and simulated. The results of the simulation showed that the FSR of comb filter is 20 GHz with a sidelobe suppression of 20 dB, the demand of FSR broadening is satisfied. Meanwhile, based on the FSP stacking over several ten nanometre optical spectrum, shaping filter is realized for sinusoidal signal generation. Finally, the potentials of low dimensional materials enhanced integrated filters for wide FSR (< 50 nm) are discussed in the last part.
Broadband radio frequency signal generation based on ultrafast optical pulse shaping, which is a typical microwave photonic technology, has been performed. The ultrafast optical pulse starts from our home-made mode-locked laser which has a repetition rate of ~171 MHz and spans from 1524 nm to 1593 nm. Following the MLL is an optical amplifier which makes up for the loss of the photonic link. A commercial pulse shaper is introduced to flatten and shape the spectrum of the optical pulse. However, it results in a additionally frequency cutoff. Only the C-band (5 THz optical bandwidth) is used efficiently which results in waste of spectrum resources. Then, the shaped pulses travel through a 5km long optical fiber realizing frequency-to-time mapping. The RF signal is acquired from the optical intensity profile by a high-speed photodetector at last. Several different kinds of wideband RF signals are generalized from the same hardware system, such as trigonometric and linear frequency-chirped waveform from 2 GHz up to 3 GHz, These results may be of interest to the radar and communication systems with ultraband RF signals.
Our experiment shows that monolayer molybdenum disulfide (ML MoS2) has big second-order non-linear susceptibility, which allows strong second harmonic generation. As the SHG has relations with the structure of crystal, we can identify the crystal’s orientation by detecting the angle dependency of SHG. We rotate the MoS2 and find the intensity of SHG has a period of 60° , which follows the symmetry of structure. At last, to detect the integrality of crystal, we perform two SHG mappings.
The abilities of operating and maintaining optical instruments are crucial in modern society. Besides the basic knowledge in optics, the optics courses in the National University of Defense Technology also focus on the training on handling typical optical equipment. As the link between classroom courses on applied optics and the field trips, the integrated design course of applied optics aims to give the students a better understanding on several instantly used optical equipment, such as hand-held telescope and periscope, etc. The basic concepts of optical system design are also emphasized as well. The course is arranged rightly after the classroom course of applied optics and composed of experimental and design tasks. The experimental tasks include the measurements of aberrations and major parameters of a primitive telescope, while in the design parts, the students are asked to design a Keplerian telescope. The whole course gives a deepened understandings on the concepts, assembling, and operating of telescopes. The students are also encouraged to extend their interests on other typical optical instruments.
Precise simulation of transient electrical behaviors of photodetectors under laser irradiation is becoming an increasingly concern. It not only can allow a detailed study and analysis of complex phenomena that cannot be carried out by experiments, but gives valuable information about the physical mechanisms which ultimately determine the response of the photodetectors. Finite difference numerical technique is adopted in the simulation to calculate the current response of photodetectors under pulsed laser irritation in this paper. To simulation the behaviors of photodetectors under pulsed laser irritation, the transport and trapping of carries and external circuit effects, including load resistance, junction capacitance, and parasitic capacitance, are considered. The basic equations governing the carrier behaviors are solved, including Poisson’s equation, the carrier motion equations, and the carrier continuity equations. The simulated transient carrier density and velocities are present, as well as corresponding transient electric field distributions. The behaviors of electrons and holes and its contributions to the external current response are analyzed. Then a general and brief image of the transient progress of photodetectors under pulsed laser irritation is established. How the carrier is induced, transported, and trapped and whether they make any significant contribution to the external current response are discussed. Besides, bias dependent response is also studied. Higher bias will improver the behaviors of photodetectors under pulsed laser irritation. The simulated results and theory analysis will show valuable clue for future research on the behaviors of photodetectors irradiated by pulsed laser.
In the experiments of photovoltaic detectors illuminated by CW lasers, some new mechanism has been
found, such as power saturation of photovoltage, hot carrier effect, as well as thermovoltage effect. To
investigate whether there is similar mechanism with pulsed laser irradiating, an 808nm femtosecond
pulsed laser is adopted. In the experiments, three photovoltaic infrared detectors are used, namely short
wavelength HgCdTe detector, medium wavelength HgCdTe detector and medium wavelength InSb
detector. Actually, the 808nm pulsed laser is spectral related laser for short wavelength HgCdTe
detector while spectral unrelated laser for medium wavelength HgCdTe and InSb detector. Under
various power densities, the detectors have a series of outputs. Power saturation of photovoltage is
observed. However, the characteristics of the outputs of these three detectors are quite different, even
between medium wavelength HgCdTe and InSb detector, which have the same packing method. There
are three major contributions in the paper. Firstly, explain the mechanism of power saturation of
photovoltage, mainly from hot carrier effect and the depressed ability of PN junction to separate
electrons and holes with the higher temperature induced by the laser. Secondly, compare the differences
between medium wavelength HgCdTe and InSb detector and give a qualitative analysis. Finally, the
difference of the outputs between short and medium wavelength HgCdTe detector is compared and
qualitatively analyzed, too, with the different mechanisms of interaction between infrared detectors and
spectral related or spectral unrelated laser. The experimental results and theory analysis will show
valuable clue for future research on photovoltaic detector irradiated by pulsed laser.
When the CCD camera is irradiated by laser, there will be "vertical bright line" phenomeno with the power of incident
laser to a certain extent. The "vertical bright line" can not be explained by the original theory of overflow crosstalk. In
this paper, the charge-transfer process as an entry point is analyzed, and the formation mechanism of the "vertical bright
line" is revealed.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.