Method for reducing relative intensity noise of QD laser on silicon substrate

Abstract

The invention provides a method for reducing relative intensity noise of a QD laser on a silicon substrate, and the method comprises the steps: selecting a master laser and a to-be-optimized slave laser, and employing a current pumping source as a driving source; calculating the optical injection locking area range of the slave laser according to the related parameters of the master laser and the slave laser; the injection laser power range and the injection laser wavelength range of the master laser are adjusted under different bias currents, meanwhile, the relative intensity noise of the slave laser is calculated, and finally the optimal relative intensity noise value within the optical injection locking area range of the slave laser can be obtained under different bias currents. The method has the beneficial effects that the relative intensity noise of the slave laser can be obviously reduced, and the increase of ground state noise caused by the excited state of the slave laser is inhibited, so that the QD laser on the silicon substrate can keep very low relative intensity noise within a very large current range.

Description

Technical Field

[0001] The invention relates to the field of semiconductor optoelectronic technology, and in particular to a method for reducing relative intensity noise of a QD laser on a silicon substrate. Background Art

[0002] In order to meet the interconnection needs of big data and high-bandwidth optical communications, silicon-based photonic integrated circuits (PICs) have been continuously improved in density, performance and stability. These advances have made PICs show broad prospects in various applications, such as being widely used in telecommunications and data center interconnection, quantum communication and computing, optical transceivers for data center interconnection, light detection and ranging (LIDAR), gyroscopes and sensing systems, and have advantages in size, weight, power consumption and cost. Quantum dots have low-dimensional semiconductor nanostructures. Semiconductor quantum dot (Quantum Dot, QD) lasers are the main candidates for high-performance photonic devices. They have low threshold current density, high temperature stability, near-zero linewidth enhancement factor (LEF) and reduced sensitivity to optical feedback. It is one of the best solutions for achieving high-performance photonic devices. Many applications such as microwave oscillators, optical gyroscopes and precision metrology have strict requirements on the noise of lasers and the overall system stability. For QD lasers on silicon substrates, due to the lattice mismatch and different thermal expansion coefficients between silicon materials and III-V materials, epitaxial growth of III-V materials directly on silicon substrates will produce a large number of mismatch dislocations, microcracks and reverse domain defects, resulting in high dislocation density. These dislocations act as non-radiative recombination centers in the Shockley-Read-Hall (SRH) process, affecting the performance of the laser, such as increasing the threshold current of the laser and increasing the noise. These problems need to be solved urgently. Although the discrete distribution characteristics of QD make its tolerance for epitaxial growth defects higher than that of quantum well structures, thus reducing the impact of dislocation defects on lasers to a certain extent, QDs on silicon substrates still have a defect density at least two orders of magnitude higher than that of native substrates.

[0003] Optical noise in semiconductor lasers includes relative intensity noise and frequency noise, the latter of which is usually converted into spectral line width, that is, the line width of QD lasers on silicon substrates. In reality, even without applied current modulation, random carrier and photon recombination and generation events will produce instantaneous time changes in carrier and photon density. Changes in photon density lead to changes in output power magnitude, thus generating background noise, and changes in carrier density lead to changes in output wavelength, thus creating a limited spectral line width for the laser mode.

[0004] Relative intensity noise mainly originates from the inherent optical phase and frequency fluctuations caused by spontaneous emission and carrier noise, which reduces the signal-to-noise ratio of the transmitted signal and increases the bit error rate, thus limiting the transmission rate and bandwidth of high-speed communication systems. In future optical communication systems, light sources with low relative intensity noise can achieve low bit error rate data transmission.

[0005] Although low relative intensity noise can be achieved by increasing the bias current of the laser source, this will bring unnecessary additional energy consumption. Although relative intensity noise can also be reduced through external optical feedback, the feedback intensity needs to be precisely controlled, otherwise the optical feedback will cause laser instability, chaos, coherence collapse and other states, which is very troublesome to operate. By continuously improving the production process of the laser and the composition of the laser, such as p-doping, the noise can be reduced to a certain extent, but it also increases the complexity of operation and the cost of implementation. Summary of the invention

[0006] In order to solve the problems in the prior art, the present invention provides a method for reducing the relative intensity noise of a QD laser on a silicon substrate. By using an external laser injection operation and adjusting the injected laser power and the injected laser wavelength of a master laser, the relative intensity noise of a slave laser can be significantly reduced, and the increase of the ground state noise caused by the excited state of the slave laser is suppressed, so that the QD laser on the silicon substrate can maintain a very low relative intensity noise within a large current range, solving the problems in the prior art of increasing the bias current of the laser source to reduce the high energy consumption, accurately controlling the external optical feedback intensity to reduce the difficulty of operation, and improving the laser production process. The method has a high cost.

[0007] The present invention provides a method for reducing the relative intensity noise of a QD laser on a silicon substrate, comprising:

[0008] Step 1, build an optical injection-locked semiconductor laser system, select a master laser and a slave laser to be optimized, and use a current pump source as a driving source for the master laser and the slave laser;

[0009] Step 2, calculating the optical injection locking region range of the slave laser, and calculating the optical injection locking region range of the slave laser according to the relevant parameters of the master laser and the slave laser, that is, the injection laser power range and the injection laser wavelength range in which the master laser optical injection coupling reduces the relative intensity noise of the slave laser;

[0010] Step 3, obtain the optimal relative intensity noise value within the optical injection locking region of the slave laser, adjust the injection laser power range and the injection laser wavelength range of the master laser under different bias currents, and calculate the relative intensity noise of the slave laser at the same time, and finally obtain the optimal relative intensity noise value within the optical injection locking region of the slave laser that can be obtained by repeatedly adjusting the injection laser power range and the injection laser wavelength range of the master laser under different bias currents, and the calculation method is as follows:

[0011] The simplified three-level rate equation model of the electronic structure of the dual-state silicon-based QD laser consists of a two-dimensional carrier pool RS and a quantum dot ensemble in the active region. The active region includes two energy levels, a quadruple degenerate excited state ES and a doubly degenerate ground state GS. This model assumes that there is only one quantum dot in the active region, and does not consider the non-uniform broadening of the gain distribution due to the uneven size of the dots. Electrons and holes are considered to be neutral excitons. External carriers are directly injected into the two-dimensional carrier pool RS. The carriers first move at a time The carriers are captured from the two-dimensional pool RS to the excited state ES, and then relax with a relaxation time From the excited state ES to the ground state GS, due to the thermal effect, the carrier escapes with a time Thermal relaxation from the ground state GS to the excited state ES is also based on the escape time The process of thermal relaxation from the excited state ES to the two-dimensional carrier pool RS is determined by the Fermi distribution of the quasi-equilibrium state. In addition, the carriers also emit spontaneously with a time Spontaneous recombination, non-radiative recombination of each energy level is determined by τ SRH Indicates that the model also includes the optical injection term S inj , establish the following description of the carrier number N RS，ES，GS , photon number S ES，GS and the electric field phase φ ES，GS The rate equation for kinetics,

[0012]

[0013] Where I is the bias current from the laser's current pump source, q is the electron charge, and Γ p is the optical confinement factor, υ g is the group velocity, τ p is the photon lifetime, β sp is the spontaneous emission factor, is the contribution of the ground state GS to the line width enhancement factor, are coefficients related to the contributions of excited states ES and two-dimensional carrier pool RS, respectively, and F RS，ES，GS , is Langevin noise, representing carrier, photon and phase noise sources respectively. When performing light injection operation, add optical injection terms to equations (2-4) to (2-7). At this time, the formula should be written as:

[0014]

[0015] Among them, S inj The number of photons injected into the main laser is defined as R inj =S inj / S0 is the injection ratio, S0 is the number of photons from the laser when it is free running, Δω inj is the frequency detuning amount, defined as the frequency difference between the master laser and the slave laser, i.e. Δω inj =ω master -ω slaυe , k c The coupling coefficient between the master laser and the slave laser is defined as L is the laser cavity length, R is the cavity surface reflectivity, and through small signal analysis and using Cramer's law, the final calculation formula of the laser relative intensity noise RIN is as follows:

[0016]

[0017] Among them, δS GS (ω) is the change in the number of photons in the frequency domain, S GS is the average number of photons.

[0018] The present invention is further improved, in step 1, the slave laser to be optimized is a QD laser on a silicon substrate whose relative intensity noise needs to be reduced.

[0019] The present invention is further improved, in step 1, the master laser is a single-mode tunable laser with a narrow spectral linewidth.

[0020] The present invention is further improved. In the step 2, the relevant parameters of the slave laser include the output power range, the output laser wavelength range, the optical parameters and the relative intensity noise of the slave laser, the relevant parameters of the master laser include the output power range and the output laser wavelength range, and the optical injection locking region range of the slave laser is the injection laser power range and the injection laser wavelength range for reducing the relative intensity noise of the slave laser after optical injection coupling of the master laser.

[0021] The present invention is further improved. In the step 2, the dynamic characteristics of the optical injection locked laser in which the light of the master laser is coupled to the slave laser depend on the injection ratio and the frequency detuning. The injection ratio is the ratio of the power of the master laser to the power of the slave laser when running freely under the same bias current. The frequency detuning is the frequency offset between the master laser and the slave laser. The injection ratio and the frequency detuning can be adjusted by adjusting the injection laser power range and the injection laser wavelength range of the master laser.

[0022] The present invention is further improved. In step 3, by changing the bias current I of the current pump source of the slave laser, the relative intensity noise of the injection parameter space under different bias currents is calculated, so that the optimal relative intensity noise value can be extracted within the optical injection locking region of the slave laser.

[0023] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention provides a method for reducing the relative intensity noise of a QD laser on a silicon substrate. By using an external laser injection operation and adjusting the injected laser power and injected laser wavelength of a master laser, the relative intensity noise of a slave laser can be significantly reduced, and the increase of the ground state noise caused by the excited state of the slave laser is suppressed, so that the QD laser on a silicon substrate can maintain a very low relative intensity noise within a large current range. Compared with the existing method of increasing the bias current of a laser source, the method of accurately controlling the external optical feedback intensity, and the method of improving the laser production process, the optimization effect is not only better, but also the energy consumption is low, the operation is simpler, and the implementation cost is low. The problems of high energy consumption in increasing the bias current of a laser source to reduce the relative intensity noise, high difficulty in operating the method of accurately controlling the external optical feedback intensity to reduce the relative intensity noise, and high cost in improving the laser production process in the prior art are solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the solutions in the present application or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 A flow chart of a method for reducing the relative intensity noise of a QD laser on a silicon substrate according to the present invention;

[0026] Figure 2 Schematic diagram of laser radar coherent detection;

[0027] Figure 3 is a schematic diagram of an optical injection system of the present invention;

[0028] Figure 4The figure shows the influence of the non-radiative recombination effect on the photocurrent characteristics of GS and ES of the present invention, as well as the threshold currents of GS (blue) and ES (red) under different non-radiative recombination lifetimes, and the corresponding ES-GS threshold ratio;

[0029] Figure 5 is a graph showing the variation of the relative intensity noise of GS and ES with the normalized bias current at different non-radiative recombination lifetimes;

[0030] Figure 6 A block diagram showing the principle of the optical injection locking process of the present invention;

[0031] Figure 7 A flow chart of the electronic structure and carrier dynamics of the QD laser on a silicon substrate of the present invention;

[0032] Figure 8 Graph showing the relative intensity noise of the GS and ES of the present invention as a function of the normalized bias current under free-running (blue) and optical injection locking (red) operations. DETAILED DESCRIPTION

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field of this application; the terms used in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit the application; the terms "including" and "having" in the specification and claims of this application and the above-mentioned figure descriptions and any variations thereof are intended to cover non-exclusive inclusions.

[0034] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0035] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0036] like Figure 2 As shown, an important application scenario of semiconductor lasers is in light detection and ranging systems, namely lidar. Figure 2The schematic diagram of the coherent detection method in the laser radar detection technology is shown. The arbitrary waveform generator generates a linear frequency modulation signal which is input into the laser modulation module as the modulation signal of the laser. One of the light emitted by the laser is used as the signal light and is sent from the transmitting system to the detection target, and the other is input into the mixer of the receiving system as the local oscillator signal. The laser modulation and drive module superimposes the frequency modulation signal with the laser bias current to form the driving current of the laser. The laser signal emitted by the laser changes with the frequency modulation signal. The signal returned by the detection target is mixed with the local oscillator signal through the receiving system. The mixed signal is photoelectrically converted into photocurrent by the photodetector. After a series of signal processing and analysis, the distance between the detection system and the detection target can be calculated. However, the relative intensity noise of the laser will affect the detection distance and resolution, and the laser intensity noise needs to be close to the shot noise.

[0037] Therefore, we propose that the relative intensity noise value of the laser output light can be reduced by optical injection, that is, the relative intensity noise of the QD laser on silicon substrate. Other parameters of the QD laser on silicon substrate are shown in the following table:

[0038] Table 1. Material and optical parameters of QD lasers on silicon substrate

[0039]

[0040]

[0041] like Figure 4 As shown, unlike quantum well lasers that emit light in a single ground state, a unique feature of silicon-based QD lasers is that they can emit light in both the ground state (GS) and the excited state (ES) simultaneously when the bias current is increased. In addition to single GS emission and simultaneous emission of GS and ES, silicon-based QD lasers can also emit light in a single ES. For dual-state emission quantum dot lasers, Figure 4 The variation of the photon number of GS and excited ES with the bias current under different non-radiative recombination lifetimes is described. It can be seen that after ES emission, the efficiency of GS emission decreases, and the growth of GS photon number is slower than that of single-state emission. The corresponding threshold current and ES-GS threshold ratio are shown in Figure 4 , the threshold current of GS changes from τ SRH = 125mA drops to τ when 0.5ns SRH = 52mA at 10ns, the threshold current of ES is from τ SRH = 412mA at 0.5ns drops to τ SRH=287mA at 10ns, and the corresponding threshold ratio of ES to GS increases from 3.3 to 5.4. It can be seen that for the free-running laser, the increase in defect density leads to an increase in threshold current, and the ES-GS threshold ratio decreases, making ES easier to emit light, which is undesirable in practical applications.

[0042] Figure 5 The variation of the relative intensity noise value at low frequency (1MHz) extracted from the noise spectrum under different non-radiative recombination lifetimes with the normalized threshold current is shown. For the dual-state QD laser on silicon substrate, it can be seen that in the free-running state, before the ES threshold, the relative intensity noise of GS decreases with the increase of bias current; near the ES threshold, the relative intensity noise of GS suddenly jumps by about 10dB, and the greater the defect density, the earlier the jump occurs; after the ES threshold, the relative intensity noise of the two laser states decreases with the increase of bias current, but the noise of GS increases overall relative to the single ground state emission. The presence of ES will increase the relative intensity noise of GS.

[0043] Therefore, the present invention provides a method for reducing the relative intensity noise of a QD laser on a silicon substrate. By using an external laser injection operation and adjusting the injected laser power and injected laser wavelength of a master laser, the relative intensity noise of a slave laser can be significantly reduced, and the increase in ground state noise caused by the excited state of the slave laser is suppressed, so that the QD laser on a silicon substrate can maintain a very low relative intensity noise within a large current range.

[0044] like Figure 1-8 As shown, the present invention provides a method for reducing the relative intensity noise of a QD laser on a silicon substrate, comprising:

[0045] Step 1, build an optical injection locked semiconductor laser system, select the master laser and the slave laser to be optimized, and use a current pump source as the driving source of the master laser and the slave laser; wherein the slave laser to be optimized is a QD laser on a silicon substrate whose relative intensity noise needs to be reduced, and the master laser is a single-mode tunable laser with a narrow spectral linewidth. In this embodiment, optical injection locking refers to injecting light from a single-mode master laser into a slave laser. When the frequency and phase of the slave laser are synchronized with the master laser, it is called stable injection locking. In this case, the dynamic characteristics of the slave laser can be greatly improved, such as suppressing RIN, nonlinear distortion, reducing frequency chirp, and enhancing modulation bandwidth.

[0046] Step 2, calculating the optical injection locking region range of the slave laser, and calculating the optical injection locking region range of the slave laser according to the relevant parameters of the master laser and the slave laser, that is, the injection laser power range and the injection laser wavelength range of the master laser optical injection coupling to reduce the relative intensity noise of the slave laser; wherein, the dynamic characteristics of the optical injection locking laser in the slave laser when the master laser is optically coupled to the slave laser depend on the injection ratio and the frequency detuning, the injection ratio is the ratio of the power of the master laser to the power of the slave laser when running freely under the same bias current, the frequency detuning is the frequency offset between the master laser and the slave laser, and the injection ratio and the frequency detuning can be adjusted by adjusting the injection laser power range and the injection laser wavelength range of the master laser; the relevant parameters of the slave laser include the output power range, the output laser wavelength range, optical parameters and relative intensity noise of the slave laser, the relevant parameters of the master laser include the output power range and the output laser wavelength range, the optical injection locking region range of the slave laser is the injection laser power range and the injection laser wavelength range of the master laser optical injection after coupling to reduce the relative intensity noise of the slave laser, by adjusting the injection laser power range and the injection laser wavelength range of the master laser, it is possible to obtain a series of relative intensity noise values ​​under injection conditions in the injection ratio and frequency detuning parameter space, and by analyzing the obtained noise values, it is possible to obtain the injection laser power and the injection laser wavelength range required for stable injection locking, that is, the optical injection locking region range of the slave laser is the injection laser power range and the injection laser wavelength range of the master laser optical injection after coupling to reduce the relative intensity noise of the slave laser.

[0047] Step 3, obtain the optimal relative intensity noise value within the optical injection locking region of the slave laser, adjust the injection laser power range and the injection laser wavelength range of the master laser under different bias currents, and calculate the relative intensity noise of the slave laser at the same time, and finally obtain the optimal relative intensity noise value within the optical injection locking region of the slave laser that can be obtained by repeatedly adjusting the injection laser power range and the injection laser wavelength range of the master laser under different bias currents, and the calculation method is as follows:

[0048] The simplified three-level rate equation model of the electronic structure of the dual-state silicon-based QD laser consists of a two-dimensional carrier pool RS and a quantum dot ensemble in the active region. The active region includes two energy levels, a quadruple degenerate excited state ES and a doubly degenerate ground state GS. This model assumes that there is only one quantum dot in the active region, and does not consider the non-uniform broadening of the gain distribution due to the uneven size of the dots. Electrons and holes are considered to be neutral excitons. External carriers are directly injected into the two-dimensional carrier pool RS. The carriers first move at a time The carriers are captured from the two-dimensional pool RS to the excited state ES, and then relax with a relaxation time From the excited state ES to the ground state GS, due to the thermal effect, the carrier escapes with a time Thermal relaxation from the ground state GS to the excited state ES is also based on the escape time The process of thermal relaxation from the excited state ES to the two-dimensional carrier pool RS is determined by the Fermi distribution of the quasi-equilibrium state. In addition, the carriers also emit spontaneously with a time Spontaneous recombination, non-radiative recombination of each energy level is determined by τ SRH Indicates that the model also includes the optical injection term S inj , establish the following description of the carrier number N RS，ES，GS , photon number S ES，GS and the electric field phase φ ES，GS The rate equation for kinetics,

[0049]

[0050]

[0051]

[0052] Where I is the bias current from the laser's current pump source, q is the electron charge, and Γ p is the optical confinement factor, υ g is the group velocity, τ p is the photon lifetime, β sp is the spontaneous emission factor, is the contribution of the ground state GS to the line width enhancement factor, are coefficients related to the contributions of excited states ES and two-dimensional carrier pool RS, respectively, and F RS，ES，GS , is Langevin noise, representing carrier, photon and phase noise sources respectively. When performing light injection operation, add optical injection terms to equations (2-4) to (2-7). At this time, the formula should be written as:

[0053]

[0054] Among them, S inj The number of photons injected into the main laser is defined as R inj =S inj / S0 is the injection ratio, S0 is the number of photons from the laser when it is free running, Δω inj is the frequency detuning amount, defined as the frequency difference between the master laser and the slave laser, i.e. Δω inj =ω master -ω slaυe , k c The coupling coefficient between the master laser and the slave laser is defined as L is the laser cavity length, R is the cavity surface reflectivity, and through small signal analysis and using Cramer's law, the final calculation formula of the laser relative intensity noise RIN is as follows:

[0055]

[0056] Among them, δS GS (ω) is the change in the number of photons in the frequency domain, S GS is the average number of photons. By changing the bias current I of the current pump source of the slave laser, the relative intensity noise of the injection parameter space under different bias currents is calculated, and the optimal relative intensity noise value can be extracted within the optical injection locking region of the slave laser. In this embodiment, Figure 8 As shown, the relative intensity noise of the injection parameter space under different bias currents is calculated, and the optimal noise value is extracted in the stable optical injection locking region. Figure 7 The comparison of the low-frequency relative intensity noise values ​​of GS and ES after optical injection and free-running operation under different normalized bias currents is shown. For GS, the relative intensity noise can be reduced by 10-13dB before the ES threshold after optical injection, by 22dB at the ES threshold, and by about 19dB after ES emission. Optical injection eliminates the adverse effects of ES emission on the relative intensity noise of GS, ensuring that GS has a lower relative intensity noise value even under high current operation. For ES, the relative intensity noise can be reduced by 19.6dB at 1.1 times the threshold current and by 11.5dB at 3.2 times the threshold.

[0057] As can be seen from the above, the present invention provides a method for reducing the relative intensity noise of a QD laser on a silicon substrate. By using an external laser injection operation to adjust the injected laser power and the injected laser wavelength of the master laser, the relative intensity noise of the slave laser can be significantly reduced, and the increase of the ground state noise caused by the excited state of the slave laser is suppressed, so that the QD laser on the silicon substrate can maintain a very low relative intensity noise within a large current range. Compared with the existing method of increasing the bias current of the laser source, the method of accurately controlling the external optical feedback intensity, and the method of improving the laser production process, the optimization effect is not only better, but also the energy consumption is low, the operation is simpler, and the implementation cost is low. The problems of high energy consumption in increasing the bias current of the laser source to reduce the relative intensity noise, the difficulty in operating the accurate control of the external optical feedback intensity to reduce the relative intensity noise, and the high cost of the method of improving the laser production process in the prior art are solved.

[0058] The specific implementation modes described above are preferred implementation modes of the present invention, and are not intended to limit the specific implementation scope of the present invention. The scope of the present invention includes but is not limited to the specific implementation modes, and all equivalent changes made according to the present invention are within the protection scope of the present invention.

Claims

1. A method for reducing the relative intensity noise of a QD laser on a silicon substrate, characterized in that: include: Step 1, build an optical injection-locked semiconductor laser system, select a master laser and a slave laser to be optimized, and use a current pump source as a driving source for the master laser and the slave laser; Step 2, calculating the optical injection locking region range of the slave laser, and calculating the optical injection locking region range of the slave laser according to the relevant parameters of the master laser and the slave laser, that is, the injection laser power range and the injection laser wavelength range in which the master laser optical injection coupling reduces the relative intensity noise of the slave laser; Step 3, obtain the optimal relative intensity noise value within the optical injection locking region of the slave laser, adjust the injection laser power range and the injection laser wavelength range of the master laser under different bias currents, and calculate the relative intensity noise of the slave laser at the same time, and finally obtain the optimal relative intensity noise value within the optical injection locking region of the slave laser that can be obtained by repeatedly adjusting the injection laser power range and the injection laser wavelength range of the master laser under different bias currents, and the calculation method is as follows: The simplified three-level rate equation model of the electronic structure of the dual-state silicon-based QD laser consists of a two-dimensional carrier pool RS and a quantum dot ensemble in the active region. The active region includes two energy levels, a quadruple degenerate excited state ES and a doubly degenerate ground state GS. This model assumes that there is only one quantum dot in the active region, and does not consider the non-uniform broadening of the gain distribution due to the uneven size of the dots. Electrons and holes are considered to be neutral excitons. External carriers are directly injected into the two-dimensional carrier pool RS. The carriers first move at a time The carriers are captured from the two-dimensional pool RS to the excited state ES, and then relax with a relaxation time From the excited state ES to the ground state GS, due to the thermal effect, the carrier escapes with a time Thermal relaxation from the ground state GS to the excited state ES is also based on the escape time The process of thermal relaxation from the excited state ES to the two-dimensional carrier pool RS is determined by the Fermi distribution of the quasi-equilibrium state. In addition, the carriers also emit spontaneously with a time Spontaneous recombination, non-radiative recombination of each energy level is determined by τ SRH Indicates that the model also includes the optical injection term S inj , establish the following description of the carrier number N RS，ES，GS , photon number S ES，GS and the electric field phase φ ES，GS The rate equation for kinetics, Where I is the bias current from the laser's current pump source, q is the electron charge, and Γ p is the optical confinement factor, v g is the group velocity, τ p is the photon lifetime, β sp is the spontaneous emission factor, is the contribution of the ground state GS to the line width enhancement factor, are coefficients related to the contributions of excited states ES and two-dimensional carrier pool RS, respectively, and F RS，ES，GS , F SES，GS , F φES，GS is Langevin noise, representing carrier, photon and phase noise sources respectively. When performing light injection operation, add optical injection terms to equations (2-4) to (2-7). At this time, the formula should be written as: Among them, S inj The number of photons injected into the main laser is defined as R inj =S inj / S0 is the injection ratio, S0 is the number of photons from the laser when it is free running, Δω inj is the frequency detuning amount, defined as the frequency difference between the master laser and the slave laser, i.e. Δω inj =ω master -ω slave , k c The coupling coefficient between the master laser and the slave laser is defined as L is the laser cavity length, R is the cavity surface reflectivity, and through small signal analysis and using Cramer's law, the final calculation formula from the laser's relative intensity noise RIN is as follows: Among them, δS GS (ω) is the change in the number of photons in the frequency domain, S GS is the average number of photons.

2. The method for reducing the relative intensity noise of a QD laser on a silicon substrate according to claim 1, characterized in that: In step 1, the slave laser to be optimized is a QD laser on a silicon substrate whose relative intensity noise needs to be reduced.

3. The method for reducing the relative intensity noise of a QD laser on a silicon substrate according to claim 2, characterized in that: In step 1, the master laser is a single-mode tunable laser with a narrow spectral linewidth.

4. The method for reducing the relative intensity noise of a QD laser on a silicon substrate according to claim 3, characterized in that: In step 2, the relevant parameters of the slave laser include the output power range, the output laser wavelength range, the optical parameters and the relative intensity noise of the slave laser, the relevant parameters of the master laser include the output power range and the output laser wavelength range, and the optical injection locking region range of the slave laser is the injection laser power range and the injection laser wavelength range for reducing the relative intensity noise of the slave laser after optical injection coupling of the master laser.

5. The method for reducing the relative intensity noise of a QD laser on a silicon substrate according to claim 4, characterized in that: In step 2, the dynamic characteristics of the optical injection locked laser in which the light of the master laser is coupled to the slave laser depend on the injection ratio and the frequency detuning. The injection ratio is the ratio of the power of the master laser to the power of the slave laser when running freely under the same bias current. The frequency detuning is the frequency offset between the master laser and the slave laser. The injection ratio and the frequency detuning can be adjusted by adjusting the injection laser power range and the injection laser wavelength range of the master laser.

6. The method for reducing the relative intensity noise of a QD laser on a silicon substrate according to claim 5, characterized in that: In step 3, by changing the bias current I of the current pump source of the slave laser and calculating the relative intensity noise in the injection parameter space under different bias currents, the optimal relative intensity noise value can be extracted within the optical injection locking region of the slave laser.

Images