RESUMO
As a photonic-based microwave signal generation method, the optoelectronic oscillator (OEO) has the potential of meeting the increasing demand of practical applications for high frequency, broadband tunability and ultra-low phase noise. However, conventional OEO systems implemented with discrete optoelectronic devices have a bulky size and low reliability, which extremely limits their practical applications. In this paper, a hybrid-integrated wideband tunable OEO with low phase noise is proposed and experimentally demonstrated. The proposed hybrid integrated OEO achieves a high integration level by first integrating a laser chip with a silicon photonic chip, and then connecting the silicon photonic chip with electronic chips through wire-bonding to microstrip lines. A compact fiber ring and an yttrium iron garnet filter are also adopted for high-Q factor and frequency tuning, respectively. The integrated OEO exhibits a low phase noise of -128.04 dBc/Hz @ 10 kHz for an oscillation frequency of 10â GHz. A wideband tuning range from 3â GHz to 18â GHz is also obtained, covering the entire C, X, and Ku bands. Our work demonstrates an effective way to achieve compact high-performance OEO based on hybrid integration, and has great potential in a wide range of applications such as modern radar, wireless communication, and electronic warfare systems.
RESUMO
The nonlinear dynamic behavior of optoelectronic oscillators (OEOs), which is important for the OEO based applications, is investigated in detail by a Microwave-photonics Iterative Nonlinear Gain (MING) model in this paper. We connect the oscillating processes with the trajectories of an iterated map based on a determined nonlinear mapping relation referred to as open-loop input to output amplitude mapping relation (IOAM). The results show that the envelope dynamic is determined by the slope of IOAM at a special point called fixed point. Linear features dominate the loop if the slope is relatively large, and the nonlinear features emerge and become increasingly significant with the decreasing of the slope. Linear features of homogeneity and monotonicity are gradually lost. Furthermore, OEO is even unstable when the slope is less than a general threshold value of -1. The behavior of OEO loops with the different slope values are discussed by simulations and are experimentally confirmed. Moreover, the proposed model also applies to the OEO with an externally injected microwave signal, the bifurcation phenomena caused by injected signal are experimentally evidenced.
RESUMO
A photonics-enabled spiking timing-dependent convolutional neural network (CNN) is proposed by manipulating photonics multidimensional parameters in terms of wavelength, temporal and spatial, which breaks the traditional CNN architecture mapping from a spatially parallel to a time-dependent series structure. The proposed CNN with the application of real-time image recognition comprises a photonics convolution processor to accelerate the computing and an involved electronic full connection to execute the classification task. A timing-dependent series of matrix-matrix operations is conducted in the photonics convolution processor that can be achieved based on multidimensional multiplexing by the accumulation of carriers from an active mode-locked laser, dispersion latency induced by a dispersion compensation fiber, and wavelength spatial separation via a waveshaper. Incorporated with the electronic full connection, a photonics-enabled CNN is proven to perform a real-time recognition task on the MNIST database of handwritten digits with a prediction accuracy of 90.04%. Photonics enables conventional neural networks to accelerate machine learning and neuromorphic computing and has the potential to be widely used in information processing and computing, such as goods classification, vowel recognition, and speech identification.
Assuntos
Redes Neurais de Computação , Óptica e Fotônica , Lasers , Aprendizado de MáquinaRESUMO
Optical frequency comb (OFC), a periodic optical pulse train in time domain, has been widely employed to measure optical frequency due to its equidistant modes in the frequency domain. Here, we propose and demonstrate a novel optical spectrum analyzer for CW lasers based on stretching the OFC in a dispersive element and mapping the frequency comb into the time domain. The optical spectrum analyzer also provides instantaneous full-field (wavelength, amplitude and phase) optical characterization capability by combining with optical phase-diversity technology. Experimental results show that we successfully trace the evolution of modulated lasers with a measurement speed of â¼51 MHz (related to the pulse repetition of the OFC) and a high spectral resolution of â¼21 pm. Thanks to the use of wavelength-to-time mapping and OFC, the single channel measurement range of the proposed system can reach â¼10 nm, which breaks the restriction of the bandwidth of photodetector.
RESUMO
Optoelectronic oscillators (OEOs) have been widely investigated to generate ultra-pure single-frequency microwave signals. Here, we propose and experimentally demonstrate a dual-chirp Fourier domain mode-locked (FDML) OEO to generate dual-chip microwave waveforms. In the proposed FDML OEO, a frequency-scanning dual-passband microwave photonics filter based on phase-modulation-to-intensity-modulation conversion using an optical notch filter and two laser diodes is incorporated into the OEO cavity. Fourier domain mode-locking operation is achieved by synchronizing the scanning period of the filter to the cavity round-trip time. Tunable dual-chirp microwave waveforms with a large time-bandwidth product are generated directly from the FDML OEO cavity in the experiment, which can be used in modern radar systems to improve its range-Doppler resolution.
RESUMO
We propose and experimentally demonstrate a photonic-assisted approach to microwave frequency measurement based on frequency-to-time mapping using a Fourier domain mode-locked optoelectronic oscillator (FDML OEO) operating around oscillation threshold. A relationship between the frequency of the unknown input microwave signals and the time difference of the output pulses is established with the help of the frequency scanning capability of the FDML OEO and, thus, can be used for microwave frequency measurement. The proposed scheme is characterized as having broad bandwidth, high resolution, multiple-frequency detection capability, and tunable measurement range. Microwave frequency measurement with a measurement range up to 16 GHz and a low measurement error of 0.07 GHz is realized.
RESUMO
A photonic-based true time delay (TTD) phased array antenna (PAA) with ultra-fast angle scan is proposed and experimentally demonstrated. A tunable TTD is realized using a wavelength-swept laser and an array of dispersive elements. The key novelty of our work is the ultra-fast angle scan using an ultra-fast wavelength-swept laser source, which is constructed by a gated multi-wavelength laser (MWL) and a dispersion compensation fiber (DCF). In our experiments, a wavelength-sweep time between two adjacent wavelengths is only several nanoseconds for wavelength spacing of 2.4, and 3.2 nm. We successfully realized an ultra-fast angle scan from 0 to 43° with a step of 8.8° in 12.48 ns.
RESUMO
An X-band optically-steered phased array radar is developed to demonstrate high resolution multi-target detection. The beam forming is implemented based on wavelength-swept true time delay (TTD) technique. The beam forming system has a wide direction tuning range of ± 54 degree, low magnitude ripple of ± 0.5 dB and small delay error of 0.13 ps/nm. To further verify performance of the proposed optically-steered phased array radar, three experiments are then carried out to implement the single and multiple target detection. A linearly chirped X-band microwave signal is used as radar signal which is finally compressed at the receiver to improve the detection accuracy. The ranging resolution for multi-target detection is up to 2 cm within the measuring distance over 4 m and the azimuth angle error is less than 4 degree.
RESUMO
Convolutional neural networks are an important category of deep learning, currently facing the limitations of electrical frequency and memory access time in massive data processing. Optical computing has been demonstrated to enable significant improvements in terms of processing speeds and energy efficiency. However, most present optical computing schemes are hardly scalable since the number of optical elements typically increases quadratically with the computational matrix size. Here, a compact on-chip optical convolutional processing unit is fabricated on a low-loss silicon nitride platform to demonstrate its capability for large-scale integration. Three 2 × 2 correlated real-valued kernels are made of two multimode interference cells and four phase shifters to perform parallel convolution operations. Although the convolution kernels are interrelated, ten-class classification of handwritten digits from the MNIST database is experimentally demonstrated. The linear scalability of the proposed design with respect to computational size translates into a solid potential for large-scale integration.