A distributed feedback laser with a double-layer grating structure

Abstract

This application provides a distributed feedback laser with a double-layer grating structure, relating to the field of laser technology. The laser consists of an InGaAlAs multi-quantum-well active region, an upper sampling grating layer, and a lower phase-shifting grating layer sequentially formed on an n-type InP substrate. The upper grating layer is designed based on reconstructed equivalent chirp technology and has two asymmetric modulation regions along the resonant cavity axis with linearly varying sampling periods in opposite directions. The lower grating layer is a quarter-wavelength phase-shifting grating with a period matched to the upper seed grating. Through the synergistic effect of equivalent phase modulation and phase shift generated by the double-layer grating, the photon density distribution within the cavity can be improved, the longitudinal spatial hole-burning effect can be reduced, and the sensitivity to end-face phase fluctuations can be decreased. This achieves high single-mode power and large modulation bandwidth over a wide temperature range, making it suitable for high-speed optical communication and coarse wavelength division multiplexing transmission systems.

Description

Technical Field

[0001] This application relates to the field of laser technology, and more specifically, to a distributed feedback laser with a double-layer grating structure. Background Technology

[0002] Distributed feedback (DFB) lasers are the core light source in modern optical communication systems, and their performance directly affects transmission rate and system stability. To improve output power and slope efficiency, the industry commonly employs high-reflection (HR) and anti-reflection (AR) end-face coatings combined with high bias current driving. However, this strategy significantly exacerbates the inhomogeneity of photon distribution within the resonant cavity, inducing the Longitudinal Spatial Hole Burning (LSHB) effect. LSHB not only induces mode jumps, relaxation oscillations, and low-frequency roll-offs, but also increases frequency chirp, severely limiting the laser's high-speed modulation capability and single-mode stability.

[0003] To suppress LSHB, the traditional approach is to introduce a λ / 4 phase-shifting structure into the grating to shift the peak photon density to the center of the resonant cavity. Although this method improves the uniformity of distribution to some extent, the peak photon density remains significant under high bias conditions, and the LSHB effect is not completely suppressed. Furthermore, the fabrication of λ / 4 phase-shifting structures typically relies on expensive and inefficient processes such as electron beam lithography (EBL), resulting in low yields and high costs. While some studies have attempted to use multiphase-shifting gratings to further optimize performance, their complex design, extreme sensitivity to random phase at the end face, and extremely small manufacturing tolerances make them unsuitable for large-scale production.

[0004] The emergence of Reconstruction Equivalent Chirp (REC) technology has provided a new approach for realizing complex grating structures at low cost. This technique, by modulating the period of the sampling grating, can equivalently generate phase modulation and chirped structures using conventional ultraviolet lithography processes, avoiding the challenges of high-precision phase shift writing. However, single REC gratings still have limitations in achieving strong coupling, highly uniform photon distribution, and low end-face phase sensitivity.

[0005] Therefore, how to design a high-performance DFB laser that can completely suppress LSHB, is insensitive to the phase of end-face reflection, and is suitable for low-cost manufacturing, while ensuring high performance, has become a key technical challenge in this field. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this application provides a distributed feedback laser with a double-layer grating structure, comprising:

[0007] Substrate;

[0008] An active region is disposed on the substrate;

[0009] The upper grating layer is located on one side of the active region; and

[0010] The lower grating layer is located on the other side of the active region;

[0011] The upper grating layer is a sampling grating based on reconstructed equivalent chirp technology, which includes at least two asymmetric modulation regions with linearly varying sampling periods along the axis of the laser resonant cavity; the lower grating layer is a phase-shifting grating with at least one phase-shifting region, and the basic grating period of the lower grating layer matches the seed grating period of the upper grating layer.

[0012] Optionally, the asymmetric modulation region of the upper grating layer includes a first asymmetric periodic gradient region and a second asymmetric periodic gradient region.

[0013] The first asymmetric periodically gradient region is located on the side of the high-reflectivity end face near the laser, and its sampling period increases linearly from away from the high-reflectivity end face to near the high-reflectivity end face.

[0014] The second asymmetric periodic gradient region is located on the side close to the anti-reflection end face of the laser, and its sampling period decreases linearly from away from the anti-reflection end face to close to the anti-reflection end face.

[0015] Optionally, the distance between the center of the first asymmetric periodically gradient region and the high-reflectivity end face is a first proportion of 0.20 to 0.25 of the total length of the resonant cavity, and its length is a second proportion of 0.08 to 0.10 of the total length of the resonant cavity;

[0016] The distance between the center of the second asymmetric periodically gradient region and the high-reflectivity end face is a third proportion of 0.65 to 0.70 of the total length of the resonant cavity, and its length is a fourth proportion of 0.12 to 0.15 of the total length of the resonant cavity.

[0017] Optionally, within the first asymmetric periodic gradient region, the sampling period ranges from 3 micrometers to 4 micrometers.

[0018] Within the second asymmetric periodic gradient region, the sampling period ranges from 4 micrometers to 4.1 micrometers.

[0019] Optionally, the lower grating layer is a quarter-wavelength phase-shift grating, the phase-shift region is located at the center of the axis of the resonant cavity, and the phase shift is π / 2;

[0020] The seed grating period of the upper grating layer ranges from 195 nanometers to 205 nanometers.

[0021] Optionally, the active region adopts an InGaAlAs multi-quantum-well structure;

[0022] The substrate is an n-type InP substrate;

[0023] The cavity length of the laser resonator ranges from 150 micrometers to 300 micrometers.

[0024] Optionally, the active region comprises 8 compressive strain trap layers and 9 tensile strain barrier layers.

[0025] The thickness of the compressive strain trap layer is 6 nanometers, and the thickness of the tensile strain barrier layer is 10 nanometers.

[0026] Optionally, the laser may further include a P-type electrode, which employs a buried organic insulator structure.

[0027] Optional, including:

[0028] A low-frequency perturbation signal sequence is superimposed on the DC drive current of the laser;

[0029] Photocurrent signals are collected by a monitoring photodiode located on the back of the laser to obtain a photocurrent time series corresponding to the low-frequency perturbation signal sequence.

[0030] The photocurrent time series is sparsely decomposed based on a preset dictionary to obtain a sparse coefficient vector. The preset dictionary includes multiple response basis vectors, each of which corresponds to the monitoring photocurrent response at different optical field center positions and matches the sampling period and position parameters of the first asymmetric periodically gradient region and the second asymmetric periodically gradient region.

[0031] The index of the current response basis vector is determined based on the sparse coefficient vector;

[0032] The control quantity is calculated based on the index of the current response basis vector and the preset reference index, and the bias current and / or operating temperature of the laser are updated according to the control quantity.

[0033] Optionally, the step of performing sparse decomposition on the photocurrent time series based on a preset dictionary to obtain a sparse coefficient vector includes:

[0034] During the device calibration stage, the resonant cavity of the laser is divided into multiple position intervals along its axis. Each position interval corresponds to a different optical field center position range within the first asymmetric periodic gradient region and the second asymmetric periodic gradient region. A corresponding bias current and operating temperature combination is set within each position interval. The monitoring photocurrent time sequence corresponding to the low-frequency perturbation signal sequence is collected, and each photocurrent time sequence is normalized and stored as a response basis vector in the preset dictionary.

[0035] During the operation phase, the inner product of the current photocurrent time series and each response basis vector in the preset dictionary is calculated, and the sparse coefficient vector is solved under the condition that the response basis vectors with non-zero sparse coefficients correspond to the same position interval.

[0036] Compared with existing technologies, this application integrates two asymmetric sampling gratings and a quarter-wavelength phase-shifting grating based on reconstruction equivalent chirp (REC) technology above and below the active region. By defining the position ratio and sampling period range of the first and second asymmetric periodically varying regions, and in conjunction with the InGaAlAs multi-quantum-well active region and the buried organic insulator P-type electrode structure, a double-grating DFB laser with both high coupling strength and relatively uniform photon distribution is achieved along the cavity length. Compared with structures using only a single-layer REC grating or a single-layer λ / 4 phase-shifting grating, the longitudinal spatial hole-burning effect is reduced, the sensitivity to end-face random phase and external reflections is decreased, and the single-mode yield, output power, and high-temperature operating stability are comprehensively improved. Simultaneously, the complex grating structure can... This method is implemented using conventional InP UV holographic exposure and lithography processes, eliminating the need for electron beam lithography, which helps reduce manufacturing costs and improve batch consistency. Based on this, the driving control method proposed in this application utilizes the monitoring photocurrent time series obtained under the condition of superimposing a low-frequency perturbation signal in the DC driving current. Combined with the response dictionary corresponding to the parameters of the two asymmetric periodically changing regions mentioned above, sparse decomposition is performed to obtain the sparse coefficient vector reflecting the location interval of the optical field center. Based on this, the control quantities of bias current and / or operating temperature are calculated. Compared with the traditional APC / ATC control scheme that only relies on average power or a single harmonic component, this method is more suitable for fine adjustment of the operating point of the intracavity optical field and phase shift region of the dual-layer grating DFB laser of this application, and is applicable to wide-temperature, high-speed modulation scenarios. Attached Figure Description

[0037] Figure 1 A schematic diagram of a corrugated modulation grating with two asymmetric structures is provided in this application embodiment;

[0038] Figure 2 A schematic diagram of a distributed feedback laser with two asymmetric corrugated modulation gratings and a λ / 4 phase shift grating provided for embodiments of this application;

[0039] Figure 3 A flowchart of a drive control method provided in an embodiment of this application;

[0040] Figure 4 This is a flowchart illustrating a method for obtaining a sparse coefficient vector, as provided in an embodiment of this application. Detailed Implementation

[0041] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0042] This embodiment provides a distributed feedback (DFB) laser with a double-layer grating structure. By integrating grating layers with different modulation characteristics on both sides of the active region, it achieves coordinated control of the optical field distribution within the laser cavity. Please refer to the appendix to the specification. Figure 1 and appendix Figure 2 ,in Figure 2 The longitudinal stacked structure of the laser is shown. Figure 1 The principle of the planar unfolding structure of the upper grating is demonstrated.

[0043] Figure 1 This application provides a schematic diagram of a corrugated modulation grating with two asymmetric structures, where HR represents the high-reflectivity end face and AR represents the anti-reflectivity end face. This indicates the seed grating period of the upper grating. , These represent the distances from the high-reflectivity end face HR to the center positions of the first and second asymmetric periodic gradient regions, respectively. , These represent the lengths of the first and second asymmetric periodically gradient regions along the axial direction of the resonant cavity, respectively. This indicates the sampling period located within a flat region outside the two asymmetric periodically transitioning regions. and These are schematic values ​​of the local sampling period in the first and second asymmetric periodic gradient regions, respectively, used to represent the linearly gradual change of the sampling period along the cavity length in each region.

[0044] Figure 2This is a schematic diagram of a distributed feedback laser with two asymmetric corrugated modulation gratings and a λ / 4 phase-shift grating, provided as an embodiment of this application. In the figure: 1 is an n-type InP substrate; 2 is an InP buffer layer; 3 is a λ / 4 phase-shift grating; 4 is an active region composed of a strain trap layer and a barrier layer; 5 is an AlGaInAs confinement layer; 6 is two asymmetric corrugated modulation gratings; 7 is an InP spacer layer; 8 is an etch stop layer; 9 is an InP capping layer; and 10 is a contact layer.

[0045] like Figure 2 As shown, the distributed feedback laser of this embodiment mainly consists of a substrate 1, an active region 4 disposed on the substrate 1, and an upper grating layer 6 and a lower grating layer 3 located on the upper and lower sides of the active region 4, respectively. The substrate 1 is an n-type InP substrate. An InP buffer layer 2 is epitaxially grown between the substrate 1 and the lower grating layer 3 to improve lattice quality and heat dissipation conditions. The active region 4 is composed of alternating strain trap layers and barrier layers. An AlGaInAs confinement layer 5, two asymmetric corrugated modulation gratings 6, an InP spacer layer 7, an etch stop layer 8, an InP capping layer 9, and a contact layer 10 are sequentially disposed on top of it to form a complete waveguide and electrode structure.

[0046] Active region 4, acting as the optical gain medium, generates stimulated emission under current injection conditions. The resulting optical field is constrained vertically by the refractive index structure of confinement layer 5, spacer layer 7, and capping layer 9, while longitudinally it is simultaneously affected by feedback from both the upper and lower grating layers. The lower grating layer 3, located below active region 4 (closer to substrate 1), is a phase-shifting grating extending along the resonant cavity axis. It has a predetermined basic grating period and at least one phase-shifting region at the axial position to introduce a phase jump within the grating reflection band to select the target longitudinal mode. The upper grating layer 6, located above active region 4 (away from substrate 1), is a sampling grating designed based on reconstructed equivalent chirp (REC) technology. Figure 1 As shown, the upper grating layer 6 is microscopically composed of a period of The seed grating is constructed, and macroscopically it is controlled by the sampling period. The spatial distribution is modulated, and at least two asymmetric modulation regions with linearly varying sampling periods are included along the cavity axis to construct an equivalent refractive index distribution that varies with the cavity length above the active region. To achieve the synergistic effect of the two grating layers, the basic grating period of the lower grating layer 3 and the seed grating period of the upper grating layer 6 are... The design is such that the phase shift feedback of the lower grating and the equivalent chirp modulation of the upper grating work together to produce the resonant mode corresponding to the same Bragg wavelength.

[0047] With the above structure, the laser achieves strong mode locking by using the lower grating, while breaking the symmetrical distribution of the intracavity optical field by using the asymmetric linear gradient structure of the upper grating. This effectively flattens the photon density distribution along the cavity length while maintaining a high side-mode suppression ratio, and suppresses the longitudinal spatial hole burning effect under high injection.

[0048] Although the macroscopic architecture of the double-layer grating and the overall concept of using an asymmetric linear gradient for the upper grating have been established, in actual DFB laser design, the two ends of the resonant cavity typically have completely different optical boundary conditions. Specifically, one end is usually a high-reflectivity (HR) end face coated with a high-reflectivity film (high reflectivity, strong standing wave effect), and the other end is an anti-reflective (AR) end face coated with an anti-reflective film (extremely low reflectivity, easy light field leakage). If the periodic gradient direction of the upper grating is not properly designed, for example, if both ends use the same gradient trend, or if the gradient direction does not match the reflection characteristics of the end face, it will be unable to effectively cancel the strong reflection phase interference at the HR end, and it will also be difficult to compensate for the photon density depletion at the AR end, which may instead exacerbate mode competition.

[0049] To address the aforementioned technical issues, this application further specifies the evolutionary rules of the asymmetric modulation region within the upper grating layer. Please refer again to the appendix. Figure 1 This figure shows the sampling period in the upper grating layer. Spatial distribution logic along the cavity length. In this embodiment, the asymmetric modulation region of the upper grating layer is clearly divided into two sub-regions with specific functional orientations: a first asymmetric periodic gradient region located on one side of the resonant cavity, and a second asymmetric periodic gradient region located on the other side.

[0050] Specifically, the first asymmetric periodically gradient region is located on the side of the high-reflectivity (HR) end face near the laser, corresponding to Figure 1 Middle left Region. To address the phase sensitivity issue caused by strong reflections at the HR end face, the sampling grating in this region is designed to have specific positive chirp characteristics, i.e., its sampling period increases linearly along the direction from away from the high-reflectivity end face to near the high-reflectivity end face (from inside the cavity to the end face).

[0051] In physical implementation, this means that during photolithography mask design, the width of the light-transmitting / shielding units in this region increases stepwise with a fixed increment. This periodic gradient, which "increases towards the end face," is equivalent in the frequency domain to introducing a gradually increasing phase barrier at the HR end, which can effectively "push" the light field so that it is not overly concentrated at the HR end face, thereby reducing the interference of the random phase at the end face on the master mode.

[0052] Correspondingly, the second asymmetric periodic gradient region is disposed on the side near the anti-reflective (AR) end face of the laser, corresponding to Figure 1 Middle right Region 1. To address the issue of decreased photon density due to low reflectivity at the AR end face, the sampling period in this region adopts an evolution logic completely opposite to that in the first region. Specifically, it linearly decreases in the direction from away from the anti-reflective end face to closer to it (from inside the cavity towards the end face). This periodic gradient, decreasing towards the end face, constructs the phase boundary conditions on the other side in the frequency domain, effectively pre-compensating and shaping the optical field.

[0053] Through the aforementioned differentiated design of increasing on the left and decreasing on the right, this embodiment effectively constructs an asymmetric equivalent phase shift distribution curve in the upper grating. This curve precisely adapts to the drastically different boundary reflection conditions at the HR and AR ends, smoothing the standing wave interference at the high-reflection end using a linearly increasing region and compensating for energy leakage at the anti-reflection end using a linearly decreasing region. This synergistic mechanism significantly improves the flatness of the photon distribution within the resonant cavity through pure geometric design without introducing additional phase shifting elements.

[0054] Having clarified the gradient trend of the asymmetric modulation region in the upper grating, this application, in order to further precisely control the photon density distribution within the resonant cavity, numerically defines the specific spatial locations and length proportions of the first and second asymmetric periodically gradient regions within the resonant cavity. The selection of these geometric parameters is based on simulation analysis of the intracavity optical field transmission characteristics and boundary reflection conditions, aiming to optimize the spatial distribution of the equivalent phase modulation to maximize the flatness of the photon density distribution.

[0055] See Figure 1 The total axial length of the laser resonant cavity is defined as L, which is the distance from the high-reflection end face to the anti-reflection end face.

[0056] In specific implementation, the first asymmetric periodic gradient region (corresponding to Figure 1 middle The region is mainly used to control the light field near the high reflectivity (HR) end face.

[0057] The distance from the center of this region to the highly reflective end face is denoted as . This embodiment will The first ratio of the total length L of the resonant cavity ( The value is limited to the range of 0.20 to 0.25. This position selection allows the phase modulation to avoid strong standing wave nodes near the HR end face, while being located in the rising edge region where optical field energy begins to accumulate, thus effectively interfering with the optical field distribution.

[0058] The length of this region along the axial direction is denoted as This embodiment will The second ratio of the total length L of the resonant cavity ( The length is limited to the range of 0.08 to 0.10. This length range ensures that the slope of the linearly gradual change of the sampling period is within a moderate range, providing sufficient phase accumulation to smooth the optical field while avoiding excessive light scattering loss due to an excessively long modulation region.

[0059] In specific implementation, the second asymmetric periodic gradient region (corresponding to) Figure 1 middle The region is mainly used to control the light field distribution on the side near the anti-reflective (AR) end face.

[0060] The distance from the center of this region to the highly reflective end face is denoted as . This embodiment will The third ratio of the total length L of the resonant cavity ( The modulation region is limited to the range of 0.65 to 0.70. The modulation region is set at about 2 / 3 of the cavity length to perform phase compensation before the light field is transmitted to the AR end and attenuates, thereby increasing the photon density in this region.

[0061] The length of this region along the axial direction is denoted as This embodiment will The fourth ratio of the total length L of the resonant cavity ( The length is limited to the range of 0.12 to 0.15. Compared to the first region, the second region has a larger length ratio, which is to provide a smoother refractive index gradient at the low reflectivity AR end to match the smoother light field gradient change there.

[0062] For example, a set of specific design parameters based on a particular cavity length: setting the total length L of the laser resonant cavity to 300 micrometers.

[0063] Based on the first ratio (0.20-0.25), the center position of the first asymmetric periodic gradient region. The selected location is 60 to 75 micrometers from the HR end face;

[0064] According to the second ratio (0.08-0.10), the length of the first asymmetric periodic gradient region Selected to be 24 micrometers to 30 micrometers;

[0065] According to the third ratio (0.65-0.70), the center position of the second asymmetric periodic gradient region. The selected location is 195 to 210 micrometers from the HR end face;

[0066] According to the fourth ratio (0.12-0.15), the length of the second asymmetric periodic gradient region The selected range is 36 micrometers to 45 micrometers.

[0067] After establishing the spatial geometric layout of the asymmetric modulation region, in order to accurately quantify the equivalent phase shift introduced by the reconstruction equivalent chirp (REC) technique, this application further defines the sampling period in each region. The specific numerical range is determined by the sampling period. According to the REC technology principle, the size of the sampling period directly determines the detuning amount and phase accumulation rate of the equivalent sub-grating. Considering the drastically different reflection boundary conditions at both ends of the resonant cavity, this embodiment employs differentiated periodic numerical designs for the first and second asymmetric periodically gradient regions to achieve distinct asymmetric phase compensation.

[0068] For the first asymmetric periodically varying region located near the high reflectivity (HR) end face, this embodiment limits the sampling period to a range of 3 micrometers to 4 micrometers. Combined with the linear increasing trend described above, within this region, the sampling period gradually increases linearly from 3 micrometers to 4 micrometers along the direction from the cavity towards the HR end face.

[0069] The relatively large span of 3 to 4 micrometers allows for the introduction of a significant equivalent refractive index gradient within a short spatial distance, specifically within the 0.08-0.10 cavity length ratio mentioned above. This large gradient creates a steep phase barrier, which effectively suppresses the strong standing wave effect caused by the high reflectivity at the HR end face, forcing photon energy to flow back towards the cavity center, thus effectively solving the photon accumulation problem near the end face.

[0070] Conversely, for the second asymmetric periodically tapering region located near the antireflective (AR) end face, this embodiment limits the sampling period to a range of 4 micrometers to 4.1 micrometers. Combined with the linear decreasing trend described above, in this region, the sampling period gradually decreases linearly from approximately 4.1 micrometers to 4 micrometers along the direction from the cavity towards the AR end face. Compared to the first region, the period variation in this region is only on the order of 0.1 micrometers, with an extremely gentle rate of change. This precise numerical design is to accommodate the low reflectivity of the AR end face. Since the light field at the AR end is in a "quasi-traveling wave" state, leakage is prone to occur. Therefore, this gentle fine-tuning mechanism is needed to make minor smooth corrections to the light field distribution at the output end without compromising beam output quality or introducing additional scattering losses, thereby optimizing the far-field beam shape.

[0071] For example, if the upper grating is in a "flat region" other than the two gradient regions mentioned above, its sampling period remains constant. Micrometer.

[0072] In the first asymmetric periodic gradient region (near the HR end), the sampling period is... The formal changes, among which Using local coordinates, the endpoint value reaches 4.0 micrometers, achieving seamless integration with flat areas. This design introduces approximately [missing information - likely a measurement unit] at the HR end. Equivalent phase shift of magnitude.

[0073] In the second asymmetric periodic gradient region (near the AR end), the sampling period is... The form changes, with the starting value of 4.1 micrometers gradually transitioning to the ending value of 4.0 micrometers (or connecting with a flat region). This design introduces only about [missing information] at the AR end. A slight phase shift in magnitude.

[0074] It is understood that the above function form is only one implementation method, and those skilled in the art can select other sampling functions that satisfy the linear gradient relationship according to the target phase distribution.

[0075] In order to suppress the longitudinal spatial hole burning effect while ensuring that the laser has robust single-mode lasing capability in a specific wavelength range, this embodiment specifies the key physical parameters in the double-layer grating structure, the phase shift of the lower grating and the seed period of the upper grating.

[0076] The lower grating layer is specifically configured to be one-quarter wavelength ( Phase-shifting grating. To maximize the concentration of optical energy within the resonant cavity and to disrupt the degeneracy of the Fabry-Perot (FP) cavity modes using a phase-shifting structure, this embodiment precisely positions the phase-shifting region at the center of the laser resonant cavity axis. This centrally symmetrical layout complements the asymmetric gradient structure of the upper grating, allowing the photon density distribution to smoothly transition towards both ends while maintaining the central peak.

[0077] Specifically, although the phase shift may vary in conventional designs, in the dual-layer composite structure of this embodiment, in order to match the equivalent chirp modulation of the upper grating and optimize the complex coupling coefficient, the phase shift introduced by the lower grating layer at the center position is controlled to be... The selection of this phase shift is beneficial for forming a suitable mode gain difference at the center of the bandgap, thereby maintaining single-mode operation under high bias current, or preventing insufficient mode discrimination due to an excessively small phase shift, thus ensuring the single-mode stability of the device under high bias current.

[0078] For the upper grating layer, its modulation basis based on reconstructed equivalent chirp technology lies in the "seed grating." The physical value of the seed grating period directly determines the Bragg center wavelength range of the laser. According to the effective refractive index of the III-V group semiconductor material system (such as InGaAlAs / InP) used in this embodiment... In order to enable the laser's lasing wavelength to cover the coarse wavelength division multiplexing (CWDM) communication band, for example, the range of 1271nm to 1331nm, this embodiment will use the seed grating period of the upper grating layer ( The value range is limited to 195 nanometers to 205 nanometers.

[0079] According to the Bragg diffraction condition When the effective refractive index When it is approximately 3.2:

[0080] Select seed grating period Nanometers correspond to the approximately 1270 nanometer wavelength band;

[0081] Select seed grating period Nanometers correspond to the approximately 1330 nanometer wavelength band.

[0082] Therefore, setting the value range between 195 nm and 205 nm allows for precise coverage of the four main channels of the CWDM system through fine-tuning of the grating period. Simultaneously, combined with the matching requirement between the lower-layer grating period and the upper-layer seed grating period in the aforementioned embodiments, this parameter range ensures resonance synchronization between the upper and lower grating layers at the target operating wavelength, avoiding feedback efficiency degradation or dual-wavelength lasing problems caused by period mismatch.

[0083] Building upon the dual-layer grating structure defined in the above embodiments, this embodiment further specifies the basic material system and resonant cavity geometry to ensure that the distributed feedback laser possesses both excellent high-temperature operating characteristics and sufficient modulation bandwidth in high-speed optical communication applications. Regarding material selection, the laser in this embodiment is constructed based on an InP-based material system, with an n-type InP substrate. This n-type substrate not only serves as a lattice-matching substrate for epitaxial growth but also possesses good thermal and electrical conductivity, which is beneficial for effective heat dissipation and electrode contact.

[0084] For the core optical gain medium, this embodiment specifically limits the active region to adopt an InGaAlAs (indium gallium aluminum arsenide) multi-quantum-well structure. Compared with the traditional InGaAsP material system, InGaAlAs material has a larger band difference in the conduction band, which can provide a stronger electron quantum confinement effect, thereby significantly suppressing carrier leakage at high temperatures, increasing the characteristic temperature of the laser, and laying the physical foundation for the device to achieve stable operation without thermoelectric cooling in a wide temperature range.

[0085] In terms of geometric design, considering the combined impact of cavity length on laser modulation bandwidth, output power, and parasitic parameters, this embodiment strictly limits the cavity length of the laser resonator to 150 micrometers to 300 micrometers. This range is chosen based on a balance between the device's RC time constant and gain volume: if the cavity length is less than 150 micrometers, while it helps reduce photon lifetime, the limited gain medium volume makes it difficult to achieve high output power and results in higher thermal resistance; if the cavity length exceeds 300 micrometers, it significantly increases junction capacitance, leading to a larger RC time constant and thus limiting the device's 3dB modulation bandwidth. Therefore, a cavity length design of 150 to 300 micrometers ensures that the device maintains sufficient linear output power while meeting the requirements of high-speed direct modulation.

[0086] Based on the dual-layer grating structure described in the above embodiments, in a specific embodiment, the active region 4 adopts a multi-quantum well stacked structure to obtain higher differential gain and better temperature characteristics. Specifically, the active region 4 includes alternating layers of compressive strain wells and tensile strain barriers, wherein the compressive strain well layers are preferably made of InGaAlAs material, and the tensile strain barrier layers are preferably made of InGaAlAs or AlGaInAs material with an appropriate amount of Al composition.

[0087] The multi-quantum-well structure is configured with eight compressive strain well layers and nine tensile strain barrier layers stacked sequentially along the epitaxial growth direction. The well layers and barrier layers are coplanar in the lateral direction and form periodic quantum potential wells in the longitudinal direction. The thickness of each compressive strain well layer is 6 nm, and the thickness of each tensile strain barrier layer is 10 nm. The thicknesses of the well and barrier layers are selected to effectively confine the carrier wavefunction within the wells and form the desired emission wavelength, while ensuring the mechanical balance of strain in the overall epitaxial structure. This multi-quantum-well structure provides the dominant gain spectrum during current injection, providing optical gain for the selected Bragg wavelength of the double-layer grating structure.

[0088] Regarding the device electrode structure, in another embodiment, the distributed feedback laser further includes a P-type electrode disposed on the upper surface. To reduce the parasitic capacitance of the device and improve the high-frequency response, the P-type electrode employs a buried organic insulator structure. Specifically, after forming the active region 4 and the grating layer and capping layer 9 above it, the device is etched in a strip or ridge pattern. The trenches on both sides of the ridge structure are filled with organic insulating material, such as benzocyclobutene resin (BCB) or other organic insulators with low dielectric constants, and a buried structure flush or nearly flush with the surrounding semiconductor layer is formed by curing or reflowing. Subsequently, a P-type metal electrode strip is formed on the contact layer 10 at the top of the ridge. The electrode strip is isolated from the semiconductor regions on both sides by the organic insulator, thereby reducing the capacitive coupling between the P-side electrode and the underlying structure while maintaining the required current injection area. This buried organic insulator P-type electrode structure is suitable for the aforementioned double-grating DFB laser, which is beneficial for obtaining smaller parasitic parameters while ensuring output power, providing an electrical basis for the device under high bit-rate direct modulation conditions.

[0089] See Figure 3 The flowchart of a drive control method provided in this application embodiment includes steps S101 to S105, wherein:

[0090] S101: A low-frequency perturbation signal sequence is superimposed on the DC drive current of the laser;

[0091] S102: The photocurrent signal is collected by a monitoring photodiode located on the back of the laser to obtain a photocurrent time sequence corresponding to the low-frequency perturbation signal sequence;

[0092] S103: Perform sparse decomposition on the photocurrent time series based on a preset dictionary to obtain a sparse coefficient vector; the preset dictionary includes multiple response basis vectors, each of which corresponds to the monitoring photocurrent response at different optical field center positions and matches the sampling period and position parameters of the first asymmetric periodic gradient region and the second asymmetric periodic gradient region.

[0093] S104: Determine the index of the current response basis vector based on the sparse coefficient vector;

[0094] S105: Calculate the control quantity based on the index of the current response basis vector and the preset reference index, and update the bias current and / or operating temperature of the laser according to the control quantity.

[0095] Based on the aforementioned distributed feedback laser with a double-layer grating structure, this embodiment further provides a driving control method. This method is applicable to the laser described above, which has a first asymmetric periodic gradient region and a second asymmetric periodic gradient region. It is used to perform closed-loop adjustment of the bias current and / or operating temperature during device operation. This method utilizes a low-frequency perturbation signal superimposed on the DC driving current and the photocurrent time series output by the back-side monitoring photodiode. Combined with a pre-constructed response dictionary, sparse decomposition is performed to obtain a response index related to the center position of the intracavity optical field, and the driving control quantity is calculated accordingly.

[0096] In one embodiment, the laser is provided with a DC bias current and a data signal modulation current by a driving circuit, which has a bias channel configurable by a control unit. When the laser is set to enter normal operating mode, the control unit superimposes a low-frequency perturbation signal sequence onto the original DC bias current.

[0097] The low-frequency perturbation signal can be a periodically changing current waveform, such as a quasi-sine wave or a stepped sequence with a small amplitude. Its frequency is selected in a range much lower than the data signal modulation rate, such as several kilohertz to tens of kilohertz, so that the perturbation signal is slower than the data modulation on the time scale and will not cause significant interference to the high-speed data signal. At the same time, the perturbation amplitude is limited to a small part of the rated bias current, such as not exceeding a few percent of the bias current, to ensure that the average output power variation of the laser remains within the allowable range.

[0098] To obtain the dynamic response of the laser's optical output under low-frequency perturbations, a monitoring photodiode is placed on the back of the laser chip. This photodiode detects a portion of the optical power leaking out from the active region, and its output is connected to an analog-to-digital converter (ADC) module via a transimpedance amplifier circuit. During the control process, the control unit collects the output of the monitoring photodiode according to a preset sampling period. Several consecutive sampling points are grouped together to form a photocurrent time series corresponding to one or more perturbation periods. To improve the signal-to-noise ratio, the control unit can collect photocurrent over multiple consecutive perturbation periods and perform averaging or filtering operations on the resulting time series to form a current photocurrent time series for feature analysis.

[0099] The control unit's memory stores a dictionary of response basis vectors, each corresponding to the monitored photocurrent response at different optical field center positions. These different optical field center positions match the sampling period and position parameters of the first and second asymmetric periodically transition regions in the laser structure. For example, they can correspond to states where the optical field center is biased towards the high-reflection end, the anti-reflection end, or the geometric center of the resonant cavity. These response basis vectors can be acquired during device factory testing or calibration by changing parameters such as bias current and operating temperature, and then stored in the dictionary after normalization. During operation, the control unit compares the current photocurrent time series with the preset dictionary and performs sparse decomposition on the dictionary representation of the time series. For example, iteratively selecting a few response basis vectors with high correlation to the current time series yields the corresponding sparse coefficient vectors, retaining only a few response basis vectors with non-zero or significant coefficients in the entire dictionary.

[0100] After obtaining the sparse coefficient vector, the control unit determines the index of the current response basis vector based on the distribution of the sparse coefficients. Specifically, the response basis vector corresponding to the element with the largest amplitude in the sparse coefficient vector can be selected as the current dominant basis vector. This basis vector is pre-associated with the corresponding optical field center position label in the dictionary, and the control unit obtains the index information of the position of the optical field center in the cavity under the current operating state based on this.

[0101] For example, when the center of the light field of the laser gradually shifts towards the high-reflection end due to the increase of the bias current or the change of the ambient temperature, the corresponding photocurrent time series in the dictionary is closer to the response basis vector corresponding to the side of the light field that is biased towards the high-reflection end. At this time, the coefficients of the corresponding index in the sparse coefficient vector will dominate.

[0102] Subsequently, the control unit compares the index of the current response basis vector with a preset reference index, which represents the desired optical field center position, such as the state corresponding to the vicinity of the phase-shift region of the lower phase-shift grating. The control unit can read the corresponding control quantity from a preset control law or lookup table based on the relationship between the current index and the reference index. The control quantity can include at least one of the following: a small increase or decrease in the bias current and a correction to the operating temperature setpoint.

[0103] For example, when the current index indicates that the center of the optical field is shifted towards the high-reflection end, the control unit can select a set of bias current reduction or temperature reduction amounts according to the lookup table to guide the center of the optical field back towards the center of the cavity; when the current index indicates that the center of the optical field is shifted towards the anti-reflection end, the control amount in the opposite direction is selected. The control unit outputs the calculated control amount to the driver chip or temperature control module to update the bias current and / or operating temperature setpoints of the laser. The above process of acquiring photocurrent, performing sparse decomposition, determining the index, and updating the control amount can be repeated periodically, thereby dynamically fine-tuning its operating point during long-term laser operation, so that the driving conditions adaptively follow the optical field center position corresponding to the double-layer grating structure design.

[0104] For example, for a double-layer grating DFB laser with a cavity length of about 300 micrometers and operating in the coarse wavelength division multiplexing band, several representative operating states can be pre-selected, such as the optical field center being slightly offset towards the high reflection end, close to the middle of the cavity length, and slightly offset towards the anti-reflection end. The monitoring photocurrent time series under the superimposed low-frequency perturbation signal under each state is collected, and the corresponding response basis vectors are generated to form a dictionary.

[0105] When the ambient temperature rises from room temperature to a higher temperature, or when the gain characteristics of the device change due to module aging, the control unit performs sparse decomposition of the monitored photocurrent time series through the above-mentioned drive control method, identifies the direction of the optical field center offset in real time, and makes corresponding adjustments to the bias current or temperature setpoint. Thus, without changing the physical structure of the device, and in conjunction with the aforementioned structural design of the two asymmetric sampling gratings and phase shift gratings, the distribution of the optical field in the cavity can be finely adjusted during operation.

[0106] See Figure 4 The flowchart of a method for obtaining a sparse coefficient vector provided in this application embodiment includes steps S201 to S202, wherein:

[0107] S201: During the device calibration stage, the resonant cavity of the laser is divided into multiple position intervals along its axis. Each position interval corresponds to a different optical field center position range within the first asymmetric periodic gradient region and the second asymmetric periodic gradient region. A corresponding bias current and operating temperature combination is set within each position interval. A monitoring photocurrent time sequence corresponding to the low-frequency perturbation signal sequence is collected, and each photocurrent time sequence is normalized and stored as a response basis vector in the preset dictionary.

[0108] S202: During the operation phase, the inner product of the current photocurrent time series and each response basis vector in the preset dictionary is calculated, and the sparse coefficient vector is solved under the condition that the response basis vectors with non-zero sparse coefficients correspond to the same position interval.

[0109] Based on the above-mentioned driving control method, in order to make the sparse coefficient vector obtained by sparse decomposition have a clear physical meaning and correspond one-to-one with the position of the optical field center in the first asymmetric periodic gradient region and the second asymmetric periodic gradient region, this embodiment further defines the construction process of the preset dictionary and the decomposition process of the running stage, dividing them into two parts: the device calibration stage and the running stage.

[0110] During the device calibration stage, the double-grating DFB laser to be calibrated can be fixed on a test fixture with temperature control capabilities. The fixture is equipped with a constant temperature stage, a programmable laser driver, and a measurement system for acquiring and monitoring photocurrent. Based on the physical length of the resonant cavity and the positional arrangement of the first and second asymmetric periodic gradient regions along the cavity length, the resonant cavity can be divided into multiple positional intervals along its axis. For example, several subdivisions can be made near the first asymmetric periodic gradient region close to the high-reflection end, and several subdivisions can also be made near the second asymmetric periodic gradient region close to the anti-reflection end. One or more intermediate intervals can be made in the region between the two.

[0111] For example, for a device with a cavity length of 300 micrometers, the area near the first asymmetric periodic gradient region close to the high-reflection end can be divided into two position intervals, the middle of the cavity length can be divided into one position interval, and the area near the second asymmetric periodic gradient region close to the anti-reflection end can be divided into two position intervals, thus obtaining five position intervals. Each position interval corresponds to a state of the optical field center near that segment.

[0112] Subsequently, for each position interval, the laser needs to be manually operated in a representative state where the center of the optical field falls within that interval, in order to collect the monitoring photocurrent time series corresponding to that state. This can be achieved by combining previous optical field simulation results of the double-layer grating structure with actual L-I-V and spectral measurements, selecting a combination of bias current and chip temperature to shift the center of the optical field towards the high-reflection end, near the middle of the cavity length, and towards the anti-reflection end, respectively.

[0113] For example, under room temperature conditions, a lower bias current is selected to slightly shift the center of the optical field towards the high-reflection end. Then, while keeping the bias current constant, the chip temperature is gradually increased to gradually shift the center of the optical field towards the middle of the cavity and the position near the anti-reflection end. Alternatively, when the temperature is fixed, several representative operating points are selected by carefully scanning near different bias currents to cover the above-mentioned position ranges.

[0114] For each position range, under the corresponding combination of bias current and temperature, a low-frequency perturbation signal sequence with a small amplitude is superimposed on the DC bias current by the driving source, such as a sinusoidal or quasi-sine wave of several kilohertz to tens of thousands of hertz. At the same time, the monitoring photocurrent time series at that operating point is collected by the back-side monitoring photodiode and transimpedance amplifier circuit.

[0115] To improve the robustness of the dictionary, the above acquisition process can be repeated for a certain location interval under multiple slightly different combinations of bias current and temperature. Each time, a multi-cycle photocurrent time series is acquired. Preprocessing steps such as time alignment, mean removal, and bandpass filtering are used to eliminate DC components and high-frequency noise. Then, the multiple time series are averaged to obtain a representative photocurrent time series for that location interval. Finally, this representative time series is normalized to unify its amplitude range and time length. The normalized time series is stored as a response basis vector in a pre-defined dictionary, and this basis vector is associated with the corresponding location interval index. Repeating the above operation for all pre-divided resonant cavity location intervals yields a dictionary containing multiple response basis vectors. Each basis vector corresponds to the monitored photocurrent response pattern when the light field center is located in different intervals, given the sampling period and position parameters of the first and second asymmetric periodically transition regions.

[0116] During operation, when the control unit needs to determine the center position of the optical field and calculate control quantities based on the current monitored photocurrent time series, it first obtains the current photocurrent time series and performs the same data preprocessing as in the calibration phase, including time window truncation, mean removal, and amplitude normalization, to ensure compatibility with the response basis vectors in the dictionary in terms of time length and amplitude scale. Subsequently, the control unit performs inner product operations on the current photocurrent time series with each response basis vector in the preset dictionary, using the inner product result as a similarity index between the current time series and each basis vector. Since each response basis vector is constructed according to position intervals, the control unit can classify and statistically analyze the similarity according to position intervals after the inner product calculation. For example, for each position interval, the maximum value or weighted average of the absolute values ​​of the inner products of all basis vectors within that interval with the current time series is taken as the overall similarity between the current operating state and the typical state corresponding to that position interval.

[0117] Based on this, in order to reflect the constraint that "response basis vectors with non-zero sparse coefficients correspond to the same position interval", the control unit in this embodiment adopts a position interval grouping method when constructing sparse coefficient vectors: first, the position interval with the largest similarity index is selected and regarded as the candidate interval where the current light field center is located; then, only one or several response basis vectors with high similarity to the current time series are selected in the candidate interval and given non-zero coefficients, while the coefficients of response basis vectors in other position intervals are all set to zero.

[0118] The resulting sparse coefficient vector has only a few components corresponding to a single position interval that are non-zero or significantly non-zero across the entire dictionary dimension. This satisfies the requirement of sparsity and ensures that the response basis vectors corresponding to all non-zero coefficients represent the same type of light field center position range, thus avoiding the situation where multiple position intervals are activated simultaneously.

[0119] For example, after dividing the resonant cavity axially into five position intervals and calibrating several response basis vectors for each interval, the control unit acquires the current monitored photocurrent time series at a certain moment. After preprocessing, it performs inner product calculation with all basis vectors in the dictionary. Assuming the calculation results show that the inner product value of the basis vectors in the position interval "near the first asymmetric periodic gradient region near the high-reflection end" is generally high, while the inner product value of the basis vectors in other position intervals is significantly low, the control unit selects this position interval as the candidate interval for the current optical field center. Within this interval, it further selects several basis vectors with the highest inner product values ​​and assigns them non-zero sparse coefficients, while not assigning non-zero coefficients to basis vectors in other position intervals. The sparse coefficient vector constructed in this way appears in the dictionary space as a cluster of coefficients concentrated in the same position interval. The control unit can then use the index of the dominant basis vector in this cluster as the "index of the current response basis vector" as described in the previous embodiment, and further compare it with a preset reference index to calculate the control quantity used for fine-tuning the bias current and / or operating temperature, thus achieving closed-loop control of the operating point of the double-grating DFB laser.

[0120] Through the coordination of the above calibration and operation mechanisms, the sparse decomposition process not only has the characteristics of sparseness in the algorithm, but also maintains the physical correspondence between the optical field center position interval in the resonant cavity, which is suitable for the driving control requirements of the double-layer grating structure with two asymmetric periodic gradient regions in the actual application scenario.

[0121] Furthermore, through the combined design of the aforementioned double-layer grating structure, the active region InGaAlAs multi-quantum-well stack, and the graded refractive index confinement layer, the DFB laser of this application exhibits superior performance compared to traditional non-phase-shift grating structures in key static and dynamic indicators. At a high temperature of 70°C, the device threshold current is controlled within the range of approximately 10mA to 30mA, and a light output greater than 7.5mW can still be obtained at a bias current of 70mA, with a significantly improved slope efficiency compared to conventional structures. Regarding the 3dB modulation bandwidth, the device can achieve a bandwidth range of approximately 12GHz to 30GHz, exhibiting a high extinction ratio and sub-10 picosecond rise / fall times at a modulation rate of 40Gb / s, meeting the requirements of high-speed direct modulation applications. Within a wide temperature range of 25°C to 70°C, the device's side-mode suppression ratio remains above approximately 45dB. Moreover, after introducing the double-layer grating structure and the buried organic insulator P-type electrode, it still manages to balance high output power with low parasitic parameters, providing a parameter basis for stable single-mode operation at high bit rates.

[0122] In system-level applications, the embodiments of this application utilize the synergistic suppression of frequency chirp and longitudinal spatial apertures through two asymmetric periodically graded gratings in the upper layer and a λ / 4 phase-shifting grating in the lower layer, enabling the device to exhibit excellent transmission tolerance across the entire Coarse Wavelength Division Multiplexing (CWDM) channel range. Even with a 106Gb / s PAM4 signal transmitted over approximately 2 kilometers of single-mode fiber, a clear and interpretable eye diagram can still be obtained, with dispersion penalty remaining below approximately +0.8dB. The TDECQ index is within the requirements for 400GbE / 800GbE optical modules. Stable operation is achieved within a wide temperature range of 25°C to 70°C without the need for thermoelectric coolers, facilitating the construction of low-power, low-cost CWDM optical modules. Therefore, the dual-grating DFB laser proposed in this application, while balancing high temperature and high power, high-speed modulation, and wide-temperature single-mode stability, possesses the conditions for large-scale mass production suitable for data center interconnects, 5G fronthaul, and long-distance CWDM transmission systems.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A distributed feedback laser with a double-layer grating structure, characterized in that, include: Substrate; An active region is disposed on the substrate; The upper grating layer is located on one side of the active region; as well as The lower grating layer is located on the other side of the active region; The upper grating layer is a sampling grating based on reconstructed equivalent chirp technology, which includes at least two asymmetric modulation regions with linearly varying sampling periods along the axis of the laser resonant cavity; the lower grating layer is a phase-shifting grating with at least one phase-shifting region, and the basic grating period of the lower grating layer matches the seed grating period of the upper grating layer. The asymmetric modulation region of the upper grating layer includes a first asymmetric periodic gradient region and a second asymmetric periodic gradient region. The first asymmetric periodically gradient region is located on the side of the high-reflectivity end face near the laser, and its sampling period increases linearly from away from the high-reflectivity end face to near the high-reflectivity end face. The second asymmetric periodically gradient region is located on the side close to the anti-reflection end face of the laser, and its sampling period decreases linearly from away from the anti-reflection end face to close to the anti-reflection end face. The distance between the center of the first asymmetric periodically gradient region and the high-reflectivity end face is a first proportion of 0.20 to 0.25 of the total length of the resonant cavity, and its length is a second proportion of 0.08 to 0.10 of the total length of the resonant cavity; the distance between the center of the second asymmetric periodically gradient region and the high-reflectivity end face is a third proportion of 0.65 to 0.70 of the total length of the resonant cavity, and its length is a fourth proportion of 0.12 to 0.15 of the total length of the resonant cavity; Within the first asymmetric periodic gradient region, the sampling period ranges from 3 micrometers to 4 micrometers; within the second asymmetric periodic gradient region, the sampling period ranges from 4 micrometers to 4.1 micrometers.

2. A distributed feedback laser with a double-layer grating structure according to claim 1, characterized in that, The lower grating layer is a quarter-wavelength phase-shift grating, and the phase-shift region is located at the center of the axis of the resonant cavity, with a phase shift of π / 2. The seed grating period of the upper grating layer ranges from 195 nanometers to 205 nanometers.

3. A distributed feedback laser with a double-layer grating structure according to claim 1, characterized in that, The active region adopts an InGaAlAs multi-quantum-well structure; The substrate is an n-type InP substrate; The cavity length of the laser resonator ranges from 150 micrometers to 300 micrometers.

4. A distributed feedback laser with a double-layer grating structure according to claim 1, characterized in that, The active region comprises 8 layers of compressive strain traps and 9 layers of tensile strain barrier layers. The thickness of the compressive strain trap layer is 6 nanometers, and the thickness of the tensile strain barrier layer is 10 nanometers.

5. A distributed feedback laser with a double-layer grating structure according to claim 1, characterized in that, The laser also includes a P-type electrode, which employs a buried organic insulator structure.

6. A driving control method for a distributed feedback laser with a double-layer grating structure as described in claim 1, characterized in that, include: A low-frequency perturbation signal sequence is superimposed on the DC drive current of the laser; Photocurrent signals are collected by a monitoring photodiode located on the back of the laser to obtain a photocurrent time series corresponding to the low-frequency perturbation signal sequence. The photocurrent time series is sparsely decomposed based on a preset dictionary to obtain a sparse coefficient vector. The preset dictionary includes multiple response basis vectors, each of which corresponds to the monitoring photocurrent response at different optical field center positions and matches the sampling period and position parameters of the first asymmetric periodically gradient region and the second asymmetric periodically gradient region. The index of the current response basis vector is determined based on the sparse coefficient vector; The control quantity is calculated based on the index of the current response basis vector and the preset reference index, and the bias current and / or operating temperature of the laser are updated according to the control quantity.

7. A driving control method for a distributed feedback laser with a double-layer grating structure as described in claim 6, characterized in that, The sparse decomposition of the photocurrent time series based on a preset dictionary to obtain a sparse coefficient vector includes: During the device calibration stage, the resonant cavity of the laser is divided into multiple position intervals along its axis. Each position interval corresponds to a different optical field center position range within the first asymmetric periodic gradient region and the second asymmetric periodic gradient region. A corresponding bias current and operating temperature combination is set within each position interval. The monitoring photocurrent time sequence corresponding to the low-frequency perturbation signal sequence is collected, and each photocurrent time sequence is normalized and stored as a response basis vector in the preset dictionary. During the operation phase, the inner product of the current photocurrent time series and each response basis vector in the preset dictionary is calculated, and the sparse coefficient vector is solved under the condition that the response basis vectors with non-zero sparse coefficients correspond to the same position interval.

Images