Microwave photonic receivers are a promising candidate in breaking the bandwidth limitation of traditional radio-frequency (RF) receivers. To further balance the performance superiority with the requirements regarding size, weight, and power consumption (SWaP), the implementation of integrated microwave photonic microsystems has been considered an upgrade path. However, up to now, to the best of our knowledge, chip-scale fully integrated microwave photonic receivers have not been reported due to the limitation of material platforms. In this paper, we report a fully integrated hybrid microwave photonic receiver (FIH-MWPR) obtained by comprising the indium phosphide (InP) laser chip and the monolithic silicon-on-insulator (SOI) photonic circuit into the same substrate based on the low-coupling-loss micro-optics method. Benefiting from the integration of all optoelectronic components, the packaged FIH-MWPR exhibits a compact volume of 6cm3 and low power consumption of 1.2 W. The FIH-MWPR supports a wide operation bandwidth from 2 to 18 GHz. Furthermore, its RF-link performance to down-convert the RF signals to the intermediate frequency is experimentally characterized by measuring the link gain, the noise figure, and the spurious-free dynamic range metrics across the whole operation frequency band. Moreover, we have utilized it as a de-chirp receiver to process the broadband linear frequency-modulated (LFM) radar echo signals at different frequency bands (S-, C-, X-, and Ku-bands) and successfully demonstrated its high-resolution-ranging capability. To the best of our knowledge, this is the first realization of a chip-scale broadband fully integrated microwave photonic receiver, which is expected to be an important step in demonstrating the feasibility of all-integrated microwave photonic microsystems oriented to miniaturized application scenarios.

It is believed that neural information representation and processing relies on the neural population instead of a single neuron. In neuromorphic photonics, photonic neurons in the form of nonlinear responses have been extensively studied in single devices and temporal nodes. However, to construct a photonic neural population (PNP), the process of scaling up and massive interconnections remain challenging considering the physical complexity and response latency. Here, we propose a comb-based PNP interconnected by carrier coupling with superior scalability. Two unique properties of neural population are theoretically and experimentally demonstrated in the comb-based PNP, including nonlinear response curves and population activities coding. A classification task of three input patterns with dual radio-frequency (RF) tones is successfully implemented in a time-efficient manner, which allows the comb-based PNP to make effective use of the ultra-broad bandwidth of photonics for parallel and nonlinear processing.

We demonstrate a chip-scale scheme of Brillouin instantaneous frequency measurement (IFM) in a CMOS-compatible doped silica waveguide chip. In the chip-scale Brillouin IFM scheme, the frequency-to-power mapping process is achieved by one-shot detection without additional time averaging and implemented by lock-in amplification, which successfully detects the Brillouin gain of the doped silica waveguide chip in the time domain. A Costas frequency modulated signal ranging from 8 GHz to 9 GHz is experimentally measured, and the frequency measurement errors are maintained within 58 MHz.

微波光子学是一门融合了微波技术和光子技术的交叉学科，是研究光波和微波在媒质中的相互作用以及在光频域实现微波信号的产生、处理、传输及接收的微波光波融合系统。由于现有的微波光子系统大多由分立器件组成，在体积、功耗、稳定性、成本等方面仍有待提升，因此集成化是微波光子技术发展的必然趋势。文中探讨了微波光子集成技术面临的主要科学与技术问题，总结了该技术的发展现状和前沿研究进展，并对其未来发展前景进行了展望。

传统的信号、信息处理流程相对独立且繁琐，人工智能（AI）技术通过引入“信号变换+信息识别”的处理策略，提升了整体处理的智能化水平。然而，未来应用中信号与信息高度密集，要求更高的系统能效、更灵活的决策能力。文中提出了通过光电融合与集成技术有望实现信号与信息兼容一体的新型处理范式：利用光子与电子技术在电磁尺度、物理性质、实现方法等层面的互补优势，针对信号与信息整体进行直接处理，并且具有融合深层智能技术的潜力。回顾了光电融合形式下的新兴信号与信息处理体制，指出了光电混合集成对光电融合处理技术的重要支撑意义。

The explosive growth of data and information has motivated various emerging non-von Neumann computational approaches in the More-than-Moore era. Photonics neuromorphic computing has attracted lots of attention due to the fascinating advantages such as high speed, wide bandwidth, and massive parallelism. Here, we offer a review on the optical neural computing in our research groups at the device and system levels. The photonics neuron and photonics synapse plasticity are presented. In addition, we introduce several optical neural computing architectures and algorithms including photonic spiking neural network, photonic convolutional neural network, photonic matrix computation, photonic reservoir computing, and photonic reinforcement learning. Finally, we summarize the major challenges faced by photonic neuromorphic computing, and propose promising solutions and perspectives.

A microwave-chip-based coherent multi-frequency RF driver is developed for a channel-interleaved photonic analog-to-digital converter (PADC) system, which comprises a multi-class optical demultiplexer and supports a sampling speed of 40 GSa/s. The generated signals from the RF driver are adjustable in both amplitude and phase. We analyze the relationship between the characteristics of the generated RF driver signals and the demultiplexing performance in theory based on the optical signal-to-distortion ratio (OSDR). It is the most effective parameter to evaluate the performance of the demultiplexer in a PADC system without an electronic analog-to-digital converter. By precisely adjusting the amplitude and phase of signals, the OSDR is optimized. The results verify the compatibility between the RF driver and the PADC system.

We propose an optical tensor core (OTC) architecture for neural network training. The key computational components of the OTC are the arrayed optical dot-product units (DPUs). The homodyne-detection-based DPUs can conduct the essential computational work of neural network training, i.e., matrix-matrix multiplication. Dual-layer waveguide topology is adopted to feed data into these DPUs with ultra-low insertion loss and cross talk. Therefore, the OTC architecture allows a large-scale dot-product array and can be integrated into a photonic chip. The feasibility of the OTC and its effectiveness on neural network training are verified with numerical simulations.