We propose, analyze and demonstrate a side-coupled Fabry-Perot (F-P) resonator filter composed of two dissimilar waveguides with π phase-shifted Bragg grating. The resonance circuit is formed by the mode conversion between the fundamental transverse electric (TE ₀ ) modes of the two waveguides. It can realize a wide free spectral range (FSR) or an FSR-free response with a single strong resonance at the central wavelength of mode conversion. In addition, compared with the conventional F-P resonator, only one transverse mode in the proposed resonator can be coupled to the bus waveguide. Thus, it has some advantages including no intrinsic reflection to the bus waveguide and low resonance loss. The experimental results on silicon-on-insulator (SOI) platform have shown that the extinction ratio, full width at half maximum (FWHM) and quality factor ( Q -factor) of the proposed resonator are 52.5 dB, 0.125 nm and 1.24 × 10 ⁴ , respectively. The fabrication error and add-drop filters based on the proposed resonator are discussed then. The proposed resonator is believed to benefit the practical applications of optical filters and sensors.

A wideband wavelength-tunable 4×5 distributed feedback (DFB) semiconductor laser array based on the reconstruction-equivalent-chirp (REC) technique using a simple tuning scheme is demonstrated. It consists of 20 DFB lasers with 4×5 matrix interleaving distributions, two-level cascaded Y-branch optical combiners, and one active semiconductor optical amplifier (SOA), all in-series integrated on one chip. Unlike the traditional thermal-electric cooler (TEC)-based wavelength-tuning scheme, the tunable 4×5 REC-DFB laser array achieves a faster and broader continuous wavelength-tuning range using TaN thin-film heaters integrated on the AlN submount. By changing the injection current of the TaN resistor from 0 to 190 mA, the proposed tunable laser achieves a wavelength-tuning range of ∼2.5nm per channel and a total tuning of over 50 nm. This study opens up new avenues for realizing cost-effective and wide-tuning-range semiconductor lasers.

In this study, we developed a robust, ultra-wideband, and high-speed wavelength-swept distributed feedback (DFB) laser array with an 8×3 matrix interleaving structure with no movable or fragile optical components. This wavelength-swept laser (WSL) achieves a continuous (gap-free) wavelength sweeping range of 60 nm and a rapid sweeping speed of 82.7 kHz, marking the widest wavelength sweeping range reported to date for high-speed WSLs based on DFB laser arrays, to our knowledge. To achieve the high-precision mapping from the time domain to the frequency domain, a nonlinear wavelength and frequency variation measurement system based on dual Fabry–Perot (F-P) etalons is designed. The system accurately measures the dynamic relationship of frequency variations over time, enabling precise wavelength interrogation. The proposed WSL was applied to the fiber Bragg grating (FBG) sensor interrogation system. In the high-low temperature and strain experiments, the system performed real-time dynamic interrogation of FBGs for up to 3 h. The experimental results demonstrated good relative accuracy and excellent interrogation performance of the system. In the vibration experiment, the system achieved high-precision interrogation of FBG sensors for high-frequency sinusoidal vibrations up to 8 kHz. Furthermore, the system worked stably under strong vibrations and shocks. Thus, the proposed WSL is applicable to high-speed FBG sensing and optical coherence tomography applications.

A large-capacity ultra-weak fiber Bragg grating (UW-FBG) array interrogation system is proposed and implemented. The system utilizes time- and wavelength-division multiplexing methods, based on a large-range tunable laser source (TLS) with fast tuning capability. The TLS consists of a distributed feedback multi-wavelength laser array (DFB-MLA) with a wavelength tuning range of 40 nm, which is realized using the reconstruction-equivalent-chirp (REC) technique. An external SOA modulator is used to modulate the output light from the TLS into nanosecond pulsed light. We establish an experimental system and successfully achieve 50 UW-FBGs interrogation. We perform temperature, strain, and vibration sensing experiments with a interrogation linearity of above 0.9995 and an error of less than 4 pm. Theoretically, the system is estimated to have an interrogation capability of more than 10,000 UW-FBGs, and the minimum spacing of adjacent UW-FBGs is 0.4 m. This research has made significant advancements in the multiplexing capability and spatial resolution of the UW-FBG array interrogation technology, offering important application potential in fields such as battery pack, structural health monitoring, and aerospace.

A 16-channel electroabsorption modulated tunable multi-wavelength DFB laser array is proposed and experimentally demonstrated. The sixteen DFB lasers, each with a wavelength spacing of 100 GHz, are arranged in a 4 × 4 matrix. The wavelengths are designed using an interleaving scheme to reduce grating crosstalk. The grating is designed and fabricated using the reconstruction equivalent chirp (REC) technique to simplify the grating fabrication and ensure precise wavelength control. The electroabsorption modulator and the tunable multi-wavelength laser array share the same active layer to reduce fabrication complexity and costs. Wavelength tuning is achieved by selecting the active laser without altering the thermoelectric cooler (TEC). Consequently, 16 channels with 100 GHz spacing are obtained, each operating in good single-mode properties with SMSRs of above 45 dB. The average wavelength spacing is 106.1 GHz, and only 6.1 GHz deviated from the design. The semiconductor optical amplifier (SOA) is integrated to compensate for the frequency chirp and amplify the output power, thus enabling long-distance transmission. The output power is amplified to over 18 mW with a 150 mA current injected into the SOA. The 3-dB bandwidth of the EAM exceeds 20 GHz, producing clear 25 Gb/s back-to-back eye diagrams with mask margins over 40% for the 16 channels. With the assistance of the chirp compensation by the SOA, the eye diagrams of the 16 channels remain clear after 10 km of transmission. Due to the superior performance, the proposed device is promising for use in DWDM systems, particularly in applications requiring fast channel switching, such as optical packet switching and NG-PON2 technology.

With the explosive growth of data, tensor processing has emerged as a pivotal component of the next generation of artificial intelligence (AI) algorithms. Current photonic convolutional processors transform tensor convolutions into multi-channel general matrix multiplication (GeMM), which follows the path of electronic counterparts, leading to data replication and hardware complexity. In this study, we experimentally and theoretically demonstrate a photonic tensor processing unit (PTPU) with a single modulator, which offers a more concise approach for multi-channel standard tensor convolution processing, different from the channel-wise convolution method. By executing multi-tensor parallel computing instead of multi-channel parallel computing, PTPU can directly produce feature tensors without clock synchronization and delay compensation between multiple channels and allows more release of physical hardware resources. Furthermore, an integrated array of semiconductor optical amplifiers (SOAs) are used to be photonic synapses for programmable weight bank, demonstrating record-high precision of 9.2 bits for weights. In the proof-of-concept experiment, we extracted features from a 3-channel (RGB) image in horizontal and vertical directions, using integrated multi-wavelength photonic tensor kernels. We then built a 3D convolutional neural network to predict the presence of COVID-19 based on computer tomography (CT) scan data consisting of 64-channel tensors.