This PDF file contains the front matter associated with SPIE Proceedings Volume 12761, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Self-injection locking is a dynamic process that passively stabilizes the emission frequency of a laser through resonant optical feedback. In the conventional approach, the laser is self-injection locked to a high-Q microresonator via front facet coupling. However, the front facet power of such lasers is limited by nonlinear effects in the microresonator. In this study, we propose an alternative self-injection locking scheme using a back facet-coupled laser, where the power from the back facet is optimally tuned to avoid nonlinear effects in the microresonator. We develop a model for the proposed scheme and find the optimal states of the scheme.

We propose and demonstrate a scheme for improving ranging performance of Frequency-Modulated Continuous-Wave (FMCW) radar, in which the transmitting signal is generated by a Semiconductor Laser (SL) under intensity-modulated optical injection. Previous investigations have demonstrated that, through introducing optical injection, SLs can be driven into period-one (P1) oscillation under suitable injection parameters, and the oscillation frequency depends on the injection strength. As a result, through modulating the intensity of injection light, FMCW can be obtained. In this work, we adopt a triangular Frequency-Modulated (FM) signal with a frequency of 11.579 MHz (period of 86.362 ns) to modulate the injection strength, an original FWCM signal with bandwidth of 2.59 GHz (from 13.59 GHz to 16.18 GHz) is generated. Taking such a signal as the transmitting signal of radar, the relative error is within the range of (17.74 % to 26.36 %) for ranging the targets within 2.600 m. The large relative error is due to poor repeatability of the transmitting signal, which can be characterized by the frequency comb contrast of FMCW signal. In order to generate the promoted FMCW signal with higher repeatability, optical feedback with delayed time of 86.362 ns is further introduced into the SL under intensity-modulated optical injection. Under optimized feedback strength, the comb contrast of promoted FMCW signal can arrive at 32.35 dB. Taking the promoted FMCW signal as the transmitting signal of radar, the relative error for ranging the targets within 2.600 m is decreased into the range of (0.42 % to 4.61 %).

In this work, time-frequency transform systems reported recently by us, including short-time Fourier transform and wavelet-like transform, are introduced. A periodic fast frequency-sweep electrical signal is used to scan the Signal Under Test (SUT). As long as the sweep is fast enough, the SUT in a single sweep period can be seen as a stationary signal and its frequency in this period can be obtained by mapping it to low-frequency pulses using filtering and frequency-to-time-mapping technique. After obtaining the frequency information of the SUT in each sweep period, the time-frequency information of the SUT can be obtained by combining the frequency information in each sweep period. The proposed method converts the time-frequency analysis of broadband signals into the analysis of low-speed electrical pulses, greatly increasing the real-time performance of the system and not relying on dispersive mediums compared to existing photonics-assisted solutions. The method for improving the system performance is also discussed by introducing filter bandwidth manipulation technology. It is found that for a given sweep speed, a proper filter bandwidth can be found to minimize the width of the electrical pulses and optimize the system frequency resolution. The photonics-assisted analog time-frequency transform method introduced in this work has a broad application prospect in the efficient and real-time acquisition of two-dimensional time-frequency information of the electromagnetic spectrum.

We propose a scheme to generate wavelength-tunable broadband Optical Frequency Combs (OFCs) based on an optical injection gain-switched Weak-Resonant-Cavity Fabry-Perot Laser Diode (WRC-FPLD). Firstly, a sinusoidal signal with frequency of 1.6 GHz and power of 19 dBm is utilized to drive the WRC-FPLD into the gain-switched state. Then, external optical injection is introduced into the gain-switched WRC-FPLD for generating wavelength-tunable broadband OFCs. The experimental results demonstrate that, when the wavelength of injected light is in the middle of the two modes, the OFCs with larger bandwidth can be obtained. When the wavelength and power of the injection light are 1545.6522 nm and 3.397 μW, respectively, the maximum bandwidth of generated OFC can arrive at 76.8 GHz (49 comb lines), and the single sideband phase noise for the fundamental frequency of the beat signal is as low as -125.5 dBc/Hz@10 kHz. The coherence of comb lines increases with the decrease of modulation frequency. Through varying the wavelength of injection light and selecting the matched injection power, the central wavelength of OFC can be adjusted within the range of (1525 nm, 1565 nm).

Thin-Film Lithium Niobate (TFLN) is a promising platform for optical communications, microwave photonics, and quantum technologies. Compared with conventional bulk materials, the TFLN platform has received widespread attention for its ability to achieve compact and high-performance integrated photonic components. In this work, we proposed a photonic-assisted multi-format microwave signal generator based on a TFLN Mach-Zehnder Modulator (MZM). It is a system with the capacity to generate and switch between a variety of microwave waveforms that are needed for optical communication and radars. The proposed microwave photonic link features a simple architecture, including only a laser, an MZM, and a photodetector (PD). Different waveforms can be obtained by appropriately setting the driven signal and Direct Current (DC) bias of the MZM, such as Phase-Shift Keying (PSK) signal, Amplitude-Shift Keying (ASK) signal, and dual-chirp microwave signal. The proposed signal generator offers a broad operating frequency range since no optical or electrical filters are involved. In addition, no extra optical processing is required which guarantees the simplicity of the generator.

The refractive index of a material is one of the most important optical parameters. In this paper, we propose the method of Self-Mixing Interferometry (SMI) to measure the refractive index of materials. SMI is superior to other laser interferometry methods because of its characteristics of simplicity and compactness. However, SMI signals are not easy to be analyzed due to the low signal-to-noise ratio and the loss of phase information. Based on the advantages of Convolutional Neural Network (CNN), in this work, we propose a scheme to reconstruct the refractive index of materials from SMI signals based on CNN. With the injection current to the laser being driven by a sawtooth wave, we first obtain different SMI signals by letting the light passing through materials with different refractive indexes under the condition of known material thickness, and then train CNN with SMI signals. The trained network is then used to estimate the refractive indexes of materials. The results show that the method is noise-proof and has high adaptability to the measurement under different conditions.

Self-Mixing Interference (SMI) is promising high-precision measurement technology with advantage of compact structure, low implementation cost and high measurement resolution, which has been used in various applications in laboratory and engineering fields. However, measurement performance of an SMI sensor can be significantly affected by noises. In this paper we propose a solution based on the U-net, a popular deep learning scheme, to remove noises from SMI signals. U-net based deep learning is used to learn noise patterns and inherent levels from large sample data, and finally to denoise the SMI signals. Our proposed method can perform end-to-end neural network model training and directly process the original waveform. The results show that this method can effectively improve the signal-to-noise ratio of SMI signals. It is believed that this unified and precise method is able to lead to enhancement of performance of SMI laser sensors operating under noisy practical engineering conditions.

Microwave photonic technology has advantages of large bandwidth, high frequency, flexible reconfigurability and immunity to electromagnetic interference, which is widely used for the generation and processing of microwave signals. In this paper, we review our recent works about microwave signal generation and processing based on optical domain controlling, including photonic generation of multi-band and multi-format microwave waveforms, and microwave photonic temperature interrogation based on temperature-to-time mapping. The proposed works have great potential in practical applications, such as radar, measurement as well as sensing.

Multiple-wavelength laser arrays play important roles in various applications, such as optical communications, optical interconnections, as well as light detection and ranging. A 64-channel laser array with a wavelength grid of 0.8 nm at C+L band is demonstrated. The lasing modes in InP-based multiple quantum wells are built based on lateral α-Si Reconstruction Equivalent Chirp (REC) gratings. By introducing an equivalent λ/4 phase-shift in each lateral α-Si REC grating, the mode degeneracy is eliminated. The equivalent λ/4 phase-shift region has no influences on the zero-order peak but generates a sharp peak in the first order peak, selecting the lasing wavelength. When the seed grating period in lateral α-Si REC grating is fixed at 250 nm and the sampling periods are changed from 6.356 μm to 39.162 μm for 64 channels, the lateral α-Si REC gratings successfully select 64 wavelengths in C+L band at 0.8 nm wavelength interval. Utilizing the sampling periods with several hundred nanometers in the first order resonant peak, we can adjust a sequence of REC gratings more accurate than the seed grating periods with several nanometers in the zero-order resonant peak. The laser in 64-channel laser array with 0.8 nm channel spacing has a threshold of 42 mA, and output power of 74.5 mW. Our work proposes a novel method of multiple-wavelength laser arrays for hetero-integration, which could provide a potential way for the development of Wavelength Division Multiplexing (WDM) system, optical interconnection inside the data center, and photonic switching.

The high reliability and efficiency of high-power conduction cooled annular diode laser stack are critical to the side pump solid-state laser head in a DPSSL system. To obtain the higher reliability and efficiency, a high-power conduction cooled annular diode laser stack packaged by AuSn hard solder has been presented. The CTE-matched wedge-shaped submounts are designed and applied in bonding GaAs-based diode laser bars with the cavity length of 1.5 mm on a conduction cooled annular heatsink. The mechanical structural design and thermal design are conducted to evaluate the capability of the annular packaging. The bar bonding process is optimized to reduce the thermal stress and improve the spectral performances of this package. After optimizing the multiple bar bonding process, a series of 808nm QCW ⪆2000W annular diode laser stacks with a narrow spectral width are achieved, which has the average FWHM and FW90%Energy value of approximately 2.6 nm and approximately 3.6 nm at 65 °C, respectively. Also, the FW90% Energy value at 65 °C is significantly reduced from 8.03 nm to 3.84 nm. Of particular importance is the elimination of the left shoulder of the spectral profile after optimizing the multiple bar bonding process.

The investigation of the nonlinear dynamics of optically pumped microresonators has attracted extensive research due to their wide applications such as spectroscopy, optical communications and distance ranging. Depending on the operation conditions, e.g., the strength of the pumped light into the microresonator, the system may exhibit several dynamical behaviors including solitons, breathers and chaos. Among these behaviors, chaos has received increasing attention in recent years for its potential use in Random Modulation Continuous Wave (RMCW) ranging and random number generation. In this paper, a comprehensive investigation on the chaotic regime of the optically pumped microresonator is performed, giving particular attention to the role played by the system’s key parameters namely pump strength and frequency detuning. Based on the Lugiato-Lefever Equations (LLEs), numerical evaluations are performed in which the route to chaos is observed by using bifurcation diagrams and the onset chaos is determined by the maximum Lyapunov exponent. The results presented in this paper provide practical guidelines for the generation of chaos in optically pumped microresonators, which are useful for applications for example random number generation.

Energy-efficient Optical Frequency Combs (OFCs) in the form of the Dissipative Kerr Soliton (DKS) pulses can be generated in the on-chip microresonators with anomalous Group Velocity Dispersion (GVD) pumped by gain switched (GS) Semiconductor Lasers (SLs). However, the achievement of anomalous GVD microresonators needs delicate designs and specific fabrication techniques. In this paper, numerical simulations have been performed for OFCs generation using a GS SL to drive a microresonator with normal GVD which can be readily made in most of the CMOS-compatible platforms. With the aid of rate equations and Lugiato Lefever Equations (LLEs), we numerically demonstrate optical pulses with a flattop square, referred to as platicons, can be generated and their characteristics are influenced by several key system parameters, namely the SL modulation frequency, GVD coefficient and frequency detuning between the SL and microresonator. Moreover, the stability region of the platicon is obtained. Our findings provide a useful guidance and easy way for the generation of energy-efficient chip scale platicons.

We propose and experimentally demonstrate a wideband multi-channel chaotic source using a Weak Resonator Cavity Fabry-Perot Laser Diode (WRC-FPLD) with Self-Phase Modulated Optical Feedback (SPMOF), in which, phase modulation is introduced into the external optical feedback loop to improve the bandwidth of the generated multi-channel chaotic signals. The experimental results show that under appropriate optical feedback intensity, WRC-FPLD with SPMOF can generate wideband multi-channel chaotic signals, when the feedback intensity is in the range of -45 dB to -15 dB, the lasing modes in the range of 1530 nm to 1570 nm can be simultaneously driven into chaos state. In addition, the Time Delay Signature (TDS) characteristics of the generated multi-channel chaotic signals are also investigated. To highlight the advantages of the proposed scheme, we also conducted comparative experiments on the conventional optical feedback scheme without phase modulation. Using SPMOF scheme, the obtained bandwidth of multi-channel chaotic signals is improved obviously, the standard bandwidth reaches 11.5 GHz, the TDS of chaotic signal is suppressed to an indistinguishable level. Relative to traditional Fabry-Perot laser diode, WRC-FPLD used in the experiment has a smaller front reflectance, about 1/90 of the rear reflectance, and a longer cavity, so it can excite more lasing modes. The proposed scheme can generate wideband multi-channel chaotic signals with a small mode spacing, which is an ideal light source for the chaotic optical communication system using wavelength division multiplexing technology.

Based on two Response Vertical-Cavity Surface-Emitting Lasers (R-VCSELs) subject to identical chaos optical injection with phase modulation and dispersion compensation, we propose and numerically investigate a high-speed bidirectional chaotic secure communication system. The driving VCSEL (D-VCSEL) is used to generate the injected chaotic signals to two R-VCSELs. After introducing the phase modulation and dispersion compensation into injection path, the bandwidth of chaotic carrier from the linear polarization components of R-VCSEL1 and R-VCSEL2 can be extended to 45GHz, and corresponding Time Delay Signatures (TDSs) can be suppressed to about 0.15. Moreover, high-quality chaos synchronization between corresponding polarization components of two R-VCSELs but very low correlation between the D-VCSEL and any R-VCSEL can be achieved. On this basis, the bidirectional dual-channel information transmission of 30 Gbit/s over 100 km fiber link is successfully realized.

A typical self-mode-locked vertical-external-cavity surface-emitting laser must operate at the edge of the stable region of the resonant cavity. Its minimal pump spot on the gain chip is used as a soft aperture. By comparing with a Continuous-Wave (CW) laser, the pulsed laser is focused more tightly on the gain chip due to the Kerr-lens effect. This mitigates cavity loss of the pulsed laser in comparison with the CW laser, such that successive mode-locking can be initiated. The disadvantage of the above method is that the relatively small pump spot, producing relatively large thermal effect, limits the output power of the laser. To address this issue, we propose another method with the work point of the laser moved slightly from the edge of the stable region and the pump spot moderately extended, with a spot of the pulsed laser on the gain chip that could be smaller or larger than that of the CW laser. We achieve stable self-mode-locking with a record average output power of 8.18 W in a V-type resonator, limited by the applied pump power. The pulse repetition rate and width are 0.71 GHz and 1.92 ps, respectively, and the corresponding peak power is 5.6 kW.

Heterodyne Optical Phase-Locked Loop (HOPLL) is an important technology for laser frequency stabilization. In this work, a digital HOPLL system for realizing frequency offset locking of a Distributed Feedback Semiconductor Laser (DFB-SL) is designed and implemented. The system is composed of a Frequency Synthesizer (FS), a passive third-order loop filter and an adjustable gain module. The effectiveness of the system is tested by measuring the frequency offset locking of DFB-SL. The results show that the designed system provides a cost-effective, sensitive, and reliable way to lock the DFB-SL with suitable parameters, and the drift caused by environmental variations can be suppressed effectively.

Distributed Feedback (DFB) semiconductor lasers with low Relative Intensity Noise (RIN) are in demand for high-power and narrow-linewidth applications. However, there is a lack of research on the compatibility of these features, together with RIN degradation at high temperature. In this paper, the RIN characteristics of InGaAsP multi-quantum-well DFB lasers are studied through theoretical calculation and numerical investigation, the results of which are very close. Based on numerical simulation, the epitaxy layers and optical cavity structures of DFB lasers are optimized to improve the RIN performance. The simulation results show that a high-power laser with an output power up to 400 mW and a narrow-linewidth laser with a linewidth below 300 kHz can obtain a peak RIN below -166 dB/Hz and -160 dB/Hz from 0.1 to 20 GHz, respectively, meeting the requirements of light sources for microwave photonics system and coherent optical transceiver system. In terms of thermal effect, buried heterostructure lasers could effectively mitigate the deterioration of RIN compared to ridge waveguide lasers due to better temperature characteristics.

In this work, we have experimentally investigated the generation of broadband Optical Frequency Comb (OFC) based on a gain-switching Vertical-Cavity Surface-Emitting Laser (VCSEL) driven by different modulation signals (including triangular-current, sinusoidal-current and square-current) under optical injection. For such a system, various modulation signals are first used to drive the VCSEL into the gain-switching state, and then an external optical injection is further introduced into the gain-switching VCSEL to generate broadband OFC. The experiment results show that the OFC output from the square current modulated VCSEL subject to optical injection has better performance compared to other two modulation signals. Driven by square signal with matched operation parameters, a high-quality OFC can be acquired, with stable comb lines, high coherence, wide bandwidth of 72.0 GHz within 10 dB amplitude variation and low single sideband phase noise at the fundamental frequency below −113.3 dBc/Hz at 10 kHz.

Self-mixing interferometry is an interferometric technique based on the self-mixing effect. Its unique structure and operating principle allow part of the light emitted by the laser to be reflected back into the laser cavity via a remote target, producing optical output power modulations and displayed as interference waveforms. The shape of these waveforms is determined by the optical feedback factor and the linewidth enhancement factor, which also affect the behavior and measurement performance of the self-mixing interferometric system. To address the need for optical measurements in the medium feedback case, this paper proposes a method to accurately estimate the optical feedback factor and linewidth enhancement factor using the non-dominated sorting genetic algorithm II (NSGA-II). The effectiveness and robustness of the method are verified by simulation and experimental results, and preliminary tests show that it can achieve an accuracy of 0.35% and 0.56% in estimating the optical feedback factor and linewidth enhancement factor. The method has potential for practical engineering applications and promotes the development of self-mixing interferometry.

This study examines the efficiency of spin hyper-polarization of xenon nuclei via spin-exchange optical pumping in an optically opaque medium of rubidium vapor with almost complete absorption of resonant radiation. A numerical model is presented that takes into account the depletion of broadband pumping in opaque medium. The study shows that the proposed method of estimating average rubidium polarization from absorption measurements of the optical pump radiation is applicable. The authors also demonstrate theoretically and experimentally the creation of a xenon polarization gradient in an optically opaque medium and propose a technique to maximize signal from gradient-polarized xenon in NMR spectroscopy.

Semiconductor laser devices have the characteristic of small size, high integration and stable requirements, and TO package red semiconductor laser are widely used in laser display application market. Due to the limitations of the electro-optical conversion efficiency of laser devices, the active region will generate over 55% of Joule waste heat under normal operation, and the traditional TO9 packaging form with Aluminum Nitride heat sink is not conducive to the heat dissipation for high power laser application. Therefore, this has become a key factor restricting the stable operation of short wavelength red semiconductor laser devices. In order to improve the heat dissipation problems, this paper developed an upgraded TO packaging type of red wavelength laser device using single-crystal Silicon Carbide (SiC) heat-sink under constant temperature condition. The experimental results show that the output power of TO packaged red laser with self-developed single-crystal SiC heat-sink is significantly higher than that of AlN heat-sink packaged device at high temperatures. The maximum electro-optical conversion efficiency of TO devices contained SiC heat sink is also improved, and the thermal resistance of the package dissipation is effectively reduced.

In the observation of CO₂ flux, high-speed and high-precision measurement of CO₂ concentration is an urgent problem to be solved, and the collection frequency of CO₂ concentration must be greater than 10Hz. The Tunable Diode Laser Absorption Spectroscopy (TDLAS) technology is widely used in various gas monitoring fields with high requirements for sensitivity, response time, and no background gas interference. In this paper, interband cascade laser at 4.26 micron is used as the system light source, and the second harmonic peak-peak value and the Root Mean Square (RMS) of sinusoidal are used to measure the CO₂ concentration. When the laser CO₂ analyzer was developed, comparative experiments were conducted using LI-7500DS and self-developed equipment. The CO₂ concentration values measured by the laser gas analyzer and the LI-7500DS gas analyzer are synchronously collected using a sampling frequency of 10 Hz. The experimental results show that the trend of CO₂ concentration measured by the two devices is consistent, and deviation of CO₂ concentration is within ±2%, which can meet the needs of carbon dioxide flux measurement.

The all-optical switching system can effectively solve problems such as transmission delay and bottleneck bandwidth. And it has significant advantages in reducing data center costs and improving its transmission characteristics. The basic idea of this study is to achieve ultra-low power consumption, ultra-low latency, and ultra-low cost switching networks through the all-optical switching system. Due to the advantages of high bandwidth and low loss, all-optical switching system are expected to replace the existing electric switching communication network in Data Center. For this system, we produced a fast tunable laser array (C-band, 16-channel, 100 GHz-space) with a switching delay of 5 ns. Each wavelength is within the error range of the channel wavelength stipulated by the Dense Wavelength Division Multiplexing (DWDM) under the ITU-T channel standard during tuning. Based on the above-mentioned laser sources, the drive control unit of the array-type tunable DFB laser was prepared, and a DWDM all-optical switching system test bed with an arrayed waveguide grating router as the core was built. Define the above laser, drive control unit and modulator as a node that can achieve arbitrary routing in the all-optical switching system mentioned above through wavelength control. A stable data switching and transmission of 20.48 Gb/s is demonstrated, in which four nodes with four different wavelengths are adopted, and full cross routes are realized.

The rapid establishment and stable maintenance of blue-green laser communication links are the conditions for the application of underwater long-distance laser communication. In the case of platform disturbance and random disturbance of seawater channels, the dynamic adaptive Acquisition Pointing and Tracking (APT) system ensures high-precision and fast alignment of laser communication links. Aiming at the problem of communication interruption caused by the interference in underwater laser communication links, combined with the characteristics of an adjustable liquid crystal optical attenuator and photoelectric tracking system, a dynamic adaptive underwater laser communication APT control system is designed and developed in this essay. The experimental results show that the blue-green laser communication system realizes the full-duplex communication function and has engineering significance.

Spectrally Efficient Frequency-Division Multiplexing (SEFDM) is a promising solution to increase communication spectral efficiency, which can pack even more sub-carriers than the orthogonal frequency-division multiplexing in a given bandwidth. However, the cost is the introduction of Inter-Carrier Interference (ICI), increasing the difficulty of signal reception and demodulation. When SEFDM signals are incorporated into a microwave photonic link, in addition to ICI, the nonlinear interference introduced by the nonlinearity of the microwave photonic link should also be considered. In this work, an iterative algorithm for microwave photonic SEFDM transmission systems is proposed to compensate for the inherent ICI of the SEFDM signal and reduce the third-order Intermodulation Distortion (IMD3) introduced by the microwave photonic transmission link. In the digital algorithm, the received 16 Quadrature-Amplitude Modulation (QAM) SEFDM signal experiences several iterations, and in each iteration, the input SEFDM signal is modified, demodulated, and Forward Error Correction (FEC) decoded into a bit sequence, which is remapped to QAM symbols to reconstruct the interference signals for canceling the distortion. Experimental results show that 16-QAM SEFDM signals with a bandwidth compression factor of 0.85 and high nonlinearity are successfully recovered from a microwave photonic link. The proposed method integrates the demodulation of SEFDM signals with the elimination of IMD3. Due to the employment of the FEC, compared with the traditional iterative ICI compensation (IIC) algorithm only for SEFDM signal demodulation, the signal demodulation capability is improved, and the improvement of signal demodulation capability also provides great help for the elimination of IMD3.