A method and system for calibrating resonant wavelength of a microring modulator

Abstract

The application discloses a resonant wavelength calibration method and system of a micro-ring modulator, and the method comprises the following steps: periodically collecting the ambient temperature of the micro-ring modulator at a first preset working frequency; performing temperature interval judgment; if the current ambient temperature is less than a first preset temperature threshold T1, maintaining the current tuning scheme of the cooperation of thermal tuning and electric tuning; if the current ambient temperature is greater than or equal to the first preset temperature threshold T1 and less than a second preset temperature threshold T2, adjusting the target thermal tuning amplitude and the target electric tuning amplitude, and performing tuning according to the adjusted target tuning amplitude; and if the current ambient temperature is greater than or equal to the second preset temperature threshold T2, switching the current tuning scheme to a pure electric tuning scheme. The application divides three working intervals of low temperature, medium temperature and high temperature through double temperature thresholds, and adopts a differentiated calibration strategy accordingly. The total tuning amount is kept constant in the whole calibration process, and the optimal balance of power consumption, tuning range and reliability under different temperatures is realized.

Description

Technical Field

[0001] This invention belongs to the field of silicon-based electronic device technology, and particularly relates to a method and system for calibrating the resonant wavelength of a micro-ring modulator. Background Technology

[0002] With the explosive growth of large-scale artificial intelligence models, high-performance computing, and data center traffic, traditional electrical interconnects are approaching their physical limits in terms of bandwidth density, transmission loss, and energy consumption. Co-packaged optics (CPO) significantly shortens the transmission path of high-speed electrical signals by integrating optical engines with computing chips such as GPUs and ASICs on a package substrate at close range. Silicon-based microring modulators, with their micron-scale compact size, extremely low drive power consumption, intrinsically high modulation bandwidth, and natural wavelength selectivity, can be used to build high-density parallel optical interconnect systems using CPO.

[0003] However, silicon has a large thermo-optic coefficient, causing the effective refractive index of the microring resonator to change significantly with temperature. In practical CPO applications, high-end computing chips typically consume hundreds of watts, resulting in severe local temperature fluctuations within the package, with transient temperature changes reaching tens of degrees Celsius. These intense thermal disturbances are directly transmitted to the silicon photonics chip, causing rapid and significant wavelength shifts in the microring resonator. Without an effective wavelength calibration and locking mechanism, the optical carrier and the microring resonator peak will be severely mismatched, leading to a deterioration in the signal extinction ratio, loss of link optical power, and ultimately, a surge in the bit error rate or even packet loss in high-speed data transmission, rendering the optical interconnect system unstable and unreliable.

[0004] Currently, the mainstream technologies for tuning and calibrating the resonant wavelength of microrings are divided into two categories: thermo-optical tuning and electro-optical tuning. Thermo-optical tuning changes the waveguide temperature by integrating heating resistors, thereby adjusting the resonant wavelength. It has the advantage of a large tuning range, but it suffers from high static power consumption and a response speed limited by the thermal time constant of the microring. Electro-optical tuning is based on the carrier dispersion effect, changing the waveguide carrier concentration through PN junctions or capacitor structures to achieve wavelength shift. It has the advantages of fast response speed and zero steady-state power consumption, but the tuning range of a single device is limited, making it difficult to cover the full temperature range drift in CPO scenarios. A single tuning mechanism cannot simultaneously meet the requirements of a large range, high precision, and low power consumption.

[0005] Chinese patent CN118138141B discloses a wavelength search and locking system for a micro-ring modulator under PAM4 modulation. The scheme uses a micro-ring modulator with integrated heating resistor as the tuning execution unit. The output optical power signal is collected by a photodetector, and the microcontroller adjusts the heater driving voltage based on the power error signal. The whole system is divided into a wavelength search stage and a steady-state locking stage, thereby realizing closed-loop feedback locking of the resonant wavelength.

[0006] Although the solution can achieve basic wavelength locking, it only uses a single thermo-optical tuning mechanism. Under steady state, it needs to continuously maintain the heating power, resulting in high static power consumption. In a parallel CPO system with hundreds of channels, the power consumption problem is amplified. At the same time, the thermal tuning response speed is limited by the thermal time constant, making it difficult to cope with the rapid transient temperature fluctuations in the CPO scenario, and it is prone to instantaneous detuning. Summary of the Invention

[0007] The purpose of this invention is to provide a resonant wavelength calibration method and system for a micro-ring modulator, which partially solves or alleviates at least one of the above-mentioned shortcomings in the prior art. By adjusting the tuning strategy according to the current actual ambient temperature, such as thermal tuning, electrical tuning, or a combination of electrical and thermal tuning, it adapts to the comprehensive requirements of wide temperature range adaptation, low power consumption operation, and high dynamic response in CPO scenarios.

[0008] To solve the aforementioned technical problems, the present invention specifically adopts the following technical solution: A first aspect of the present invention provides a method for calibrating the resonant wavelength of a micro-ring modulator, wherein the micro-ring modulator integrates a thermal tuning unit and an electrical tuning unit, which work together to compensate for the deviation between the resonant wavelength and the target operating wavelength. The calibration method includes the following steps: The ambient temperature of the micro-ring modulator is periodically collected at a first preset operating frequency. The current ambient temperature is compared with the first preset temperature threshold T1 and the second preset temperature threshold T2 to determine the range to which the current temperature belongs. If the current ambient temperature is less than the first preset temperature threshold T1, the current tuning scheme of thermal tuning and electrical tuning will be maintained. If the current ambient temperature is greater than or equal to the first preset temperature threshold T1 and less than the second preset temperature threshold T2, then the target thermal tuning amplitude of the thermal tuning unit is reduced according to the preset first reduction, and the target electrical tuning amplitude of the electrical tuning unit is increased according to the reduced target thermal tuning amplitude; and coordinated tuning is performed according to the reduced target thermal tuning amplitude and the increased target electrical tuning amplitude. If the current ambient temperature is greater than or equal to the second preset temperature threshold T2, the current tuning scheme will be switched to the pure electric tuning scheme.

[0009] Furthermore, the first preset operating frequency is less than the second preset operating frequency, which is the operating frequency of the built-in temperature control module of the micro-ring modulator.

[0010] Furthermore, it also includes an adaptive adjustment step for the operating frequency: Obtain the temperature change rate of the current cycle and compare the temperature change rate with the first preset speed threshold and the second preset speed threshold. If the rate of temperature change is less than the first preset rate threshold, the first preset operating frequency is reduced according to the second preset reduction. If the rate of temperature change is greater than or equal to the first preset speed threshold and less than the second preset speed threshold, the current first preset operating frequency is maintained unchanged. If the rate of temperature change is greater than or equal to the second preset speed threshold, the first preset operating frequency is increased by the second preset increment.

[0011] Furthermore, when the increased first preset working frequency does not exceed the threshold ratio of the second preset working frequency, and the temperature change rate exceeds the third preset speed threshold, the increase in sampling frequency is paused, and the current temperature change slope k and the fitting correlation coefficient r² are calculated by linear regression of the temperature values ​​within M cycles. If the fitted correlation coefficient r² is greater than the preset confidence threshold, the predicted temperature value for the next cycle is calculated based on the current temperature change slope k, and the temperature range is judged based on the predicted temperature value to switch the tuning scheme. If the fitted correlation coefficient r² is less than or equal to the preset threshold, the temperature range is determined directly based on the current measured ambient temperature to switch the tuning scheme.

[0012] Furthermore, before calculating the predicted temperature value for the next cycle, an acceleration verification step is also included: Calculate the acceleration of temperature change in the current cycle; If the acceleration due to temperature change is positive and greater than or equal to the preset acceleration threshold, then the temperature range judgment is performed based on the measured ambient temperature. If the acceleration due to temperature change is negative or less than the preset acceleration threshold, the predicted temperature value for the next cycle is calculated based on the current temperature change slope k, and the temperature range is determined using this predicted temperature value.

[0013] Furthermore, the total time for the thermally tuned power to decay from its current value to zero is greater than 0.5 times the thermal time constant of the micro-ring modulator, but less than 3 times the thermal time constant of the micro-ring modulator.

[0014] Furthermore, the preset first reduction utilization formula is as follows: ΔA = A1 × (T2 - Tc) / (T2 - T1) Calculate; where ΔA is the preset first reduction, A1 is the initial target thermal tuning amplitude, and Tc is the current ambient temperature.

[0015] Furthermore, after switching to the pure electric tuning scheme, if the target electric tuning amplitude of the electric tuning unit reaches the preset safety limit and the resonant wavelength still cannot converge after the duration exceeds the preset time threshold, then wavelength locking is paused, the thermal tuning power is reduced to zero, and after a preset thermal equilibrium time, recalibration is performed.

[0016] Furthermore, when it is determined that the current ambient temperature is greater than or equal to the second preset temperature threshold T2, and during the switching process of the current tuning scheme to the pure electric tuning scheme, the thermal tuning power is linearly or exponentially decayed to zero with a preset decreasing slope, while the electric tuning bias is synchronously increased with a preset increasing slope.

[0017] A second aspect of the present invention is to provide a resonant wavelength calibration system for a micro-ring modulator, comprising: An ambient temperature acquisition module is used to periodically acquire the ambient temperature of the micro-ring modulator at a first preset operating frequency. The temperature range determination module is used to compare the current ambient temperature with the first preset temperature threshold T1 and the second preset temperature threshold T2 respectively to determine the range to which the current temperature belongs. The tuning scheme switching module is used to maintain the current tuning scheme of thermal tuning and electrical tuning when the current ambient temperature is less than the first preset temperature threshold T1; and when the current ambient temperature is greater than or equal to the first preset temperature threshold T1 and less than the second preset temperature threshold T2, to reduce the target thermal tuning amplitude of the thermal tuning unit according to the preset first reduction, and to increase the target electrical tuning amplitude of the electrical tuning unit according to the reduced target thermal tuning amplitude; to perform tuning according to the reduced target thermal tuning amplitude and the increased target electrical tuning amplitude; and when the current ambient temperature is greater than or equal to the second preset temperature threshold T2, to switch the current tuning scheme to a pure electrical tuning scheme, during which the thermal tuning power is linearly or exponentially decayed to zero with a preset decreasing slope, and the electrical tuning bias is increased synchronously with a preset increasing slope.

[0018] Beneficial effects: Typically, the temperature control module built into a micro-ring modulator periodically monitors the actual operating temperature of the micro-ring modulator at a second preset operating frequency. Based on the monitored actual temperature, and combined with existing algorithms such as PID control, it controls the thermal tuning module to perform thermal tuning, bringing it to the target temperature and thus the target thermal tuning amplitude. However, micro-ring modulators are highly sensitive to temperature. Therefore, in practical applications, the thermal tuning function of the micro-ring modulator is also affected by the ambient temperature. For example, if the ambient temperature is already high, the resonant wavelength of the micro-ring will have shifted significantly. In this case, a larger thermal tuning amplitude is required to bring the resonant wavelength back to the target, which will bring additional power consumption and thermal budget pressure, and may even exceed the adjustment range of the thermal tuning, leading to locking failure. If the ambient temperature only suddenly increases but then decreases within a short period, the actual impact is not significant, and tuning can still be performed according to the current tuning strategy. Therefore, this invention automatically adjusts the tuning strategy based on actual operating conditions, such as the current ambient temperature. Specifically, The system divides the operating range into three zones—low, medium, and high—using dual temperature thresholds, and employs differentiated calibration strategies accordingly. In the low-temperature zone, thermoelectric co-tuning is maintained, fully utilizing the wide tuning range of thermal tuning to ensure wavelength coverage across the entire temperature spectrum. In the medium-temperature zone, the thermal tuning amplitude decreases linearly with temperature while simultaneously increasing the proportion of electrical tuning, achieving a smooth transition between tuning modes. In the high-temperature zone, the system switches to pure electrical tuning mode, leveraging the zero static power consumption of electrical tuning to eliminate the additional heating power consumption of thermal tuning at high temperatures. This avoids continuous heating that exacerbates localized heat accumulation on the chip, reduces the risk of thermal crosstalk between multiple channels, and improves the operational reliability of the device under high-temperature conditions. Throughout the calibration process, the total tuning amount remains constant, achieving an optimal balance between power consumption, tuning range, and reliability at different temperatures without affecting wavelength locking stability.

[0019] During the switching process to pure electric tuning in the high-temperature range, a soft switching method is adopted, in which the thermal tuning power decays linearly or exponentially according to a preset slope, and the electric tuning bias increases synchronously and equally. The total switching time is constrained within the range of 0.5 to 3 times the thermal time constant of the micro-ring. This not only matches the thermal response inertia of the micro-ring and avoids the instantaneous jump of the resonant wavelength caused by power change, thus preventing optical signal errors and transmission interruptions and ensuring the continuity and stability of the communication link, but also avoids the system response lag caused by the switching process being too long, thus balancing the switching smoothness and response speed.

[0020] Ambient temperature is dynamic, and its fluctuations may vary under different operating conditions. If ambient temperature is considered throughout the tuning process, it will inevitably increase system power consumption and algorithm complexity or computational load. On the one hand, from the data acquisition perspective, the system frequently collects a large amount of temperature data from temperature sensors. If the ambient temperature fluctuation is not large at the current stage, the frequently collected data will become a large amount of redundant data, thereby increasing system power consumption and occupying a large amount of system memory, wasting resources. If the ambient temperature fluctuation is too large at the current stage, the data collected by traditional fixed sampling frequency is insufficient to reflect the current fluctuation, thus making it difficult to achieve sufficient accuracy even if ambient temperature is considered for thermal tuning. On the other hand, from the algorithm perspective, if the ambient temperature fluctuation is not large, the large amount of redundant data collected will greatly increase the computational load if no data processing is performed, such as data filtering, which will increase the algorithm complexity. Furthermore, if the ambient temperature fluctuation is not large, frequently adjusting the thermal tuning amplitude or switching the tuning strategy will also increase the system load and power consumption. Therefore, when incorporating ambient temperature into the thermal tuning process, a trade-off must be struck between the sampling frequency (i.e., the first operating frequency) and data redundancy, as well as algorithm complexity and system power consumption. In other words, the system needs to balance tuning accuracy, response speed, and system power consumption. Therefore, in this invention, the operating frequency of temperature sampling and strategy calibration is dynamically adjusted according to the rate of change of ambient temperature: when the temperature change is gradual, the sampling frequency is reduced to decrease microprocessor computational overhead and sampling power consumption, optimizing the overall system energy efficiency; when the temperature changes rapidly, the sampling frequency is increased to enhance the monitoring density of thermal disturbances, ensuring timely response of the calibration strategy.

[0021] This invention introduces a least-squares linear regression temperature prediction mechanism in high-speed linear temperature change scenarios. By fitting multi-period temperature data, the slope of change and correlation coefficient are obtained. Under high confidence, the predicted temperature is used to perform interval judgment and strategy switching in advance, transforming the traditional ex-post feedback correction into ex-ante prediction compensation. This effectively offsets the control lag caused by the thermal time constant of thermal tuning, improves the system's response speed to rapid linear temperature fluctuations, and avoids instantaneous detuning and signal degradation caused by transient temperature changes. Simultaneously, a temperature acceleration verification step is added. By judging the trend of temperature change rate, nonlinear temperature change scenarios are identified. The prediction mechanism is disabled when the temperature changes rapidly, avoiding prediction misjudgments in nonlinear scenarios, and significantly improving the robustness of feedforward calibration and its adaptability to complex thermal environments.

[0022] In addition, this invention sets up a fault protection mechanism for wavelength loss-of-lock scenarios under extreme operating conditions. When the electrical tuning amplitude reaches the safety limit and cannot converge after a timeout, the thermal tuning power is forcibly reset. After the thermal field is balanced, the calibration process is automatically restarted. This not only avoids the system from falling into a dead loop of repeated adjustments without being able to lock, but also prevents the electrical tuning from working under overvoltage conditions for a long time and the thermal tuning from operating at high temperatures for a long time, thus effectively improving the system's fault tolerance and self-healing ability under extreme operating conditions, and ensuring the long-term reliability of the device.

[0023] This invention addresses the multi-channel parallel micro-ring array scenario by employing a time-division polling calibration scheduling mechanism. Within a single calibration cycle, only the tuning parameters of a single channel are updated, while the thermal tuning power of the remaining channels is frozen. The polling interval is set to three times the micro-ring thermal time constant, ensuring that thermal field fluctuations caused by single-channel tuning completely dissipate before executing the next channel calibration. This avoids thermal field interference and wavelength-locked oscillations caused by simultaneous tuning of multiple channels at the control level, effectively solving the industry problem of inter-channel thermal crosstalk under high-density integration. It can directly adapt to the high-density integration requirements of hundreds of parallel optical channels in co-packaged optics scenarios, improving the overall operational stability of large-scale micro-ring arrays. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. The elements or parts in the drawings are not necessarily drawn to scale. Obviously, the drawings described below are some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0025] Figure 1 This is a flowchart of Embodiment 1 of the present invention.

[0026] Figure 2 This is a structural diagram of Embodiment 3 of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0028] In this document, suffixes such as "module," "part," or "unit" used to denote elements are used only for the purpose of illustrative purposes and have no specific meaning in themselves. Therefore, "module," "part," or "unit" may be used interchangeably.

[0029] In this document, the terms "upper," "lower," "inner," "outer," "front," "rear," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the present invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0030] In this document, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, a direct connection, or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0031] In this document, "and / or" includes any and all combinations of one or more of the listed related items.

[0032] In this article, "multiple" means two or more, that is, it includes two, three, four, five, etc.

[0033] Example 1: like Figure 1 As shown, this embodiment provides a method for calibrating the resonant wavelength of a microring modulator, applied to a wavelength locking system for silicon-based microring modulators, and particularly suitable for high-density microring modulator arrays in co-packaged optical scenarios. The wavelength locking system for the microring modulator integrates a thermal tuning unit and an electrical tuning unit, which work together to compensate for the deviation between the microring resonant wavelength and the target operating wavelength.

[0034] The wavelength locking system also includes a thermistor, a microprocessor, a signal conditioning circuit, and a photoelectric detection unit. The thermistor is deployed on a silicon photonic chip surrounding the micro-ring modulator and is used to collect the ambient temperature of the micro-ring. The microprocessor is the control core of the system, used to perform temperature acquisition, range judgment, tuning parameter calculation, and control signal output. The micro-ring modulator has a built-in low-level temperature control module that uses existing algorithms such as PID control to directly drive the thermal tuning unit, making the actual thermal tuning amplitude quickly approach the target thermal tuning amplitude. The operating frequency of this built-in temperature control module is a second preset operating frequency, which is relatively high and used to achieve rapid closed-loop temperature control. This is existing technology and will not be described further here.

[0035] Specifically, the thermal tuning unit is a titanium nitride heating resistor integrated around the microring resonant waveguide, or a silicon-based heating resistor formed by ion doping. Its working principle involves applying a driving voltage to the heating resistor, which generates Joule heating, changing the local temperature of the microring waveguide. The thermo-optical effect of silicon is used to increase the effective refractive index of the waveguide, causing the microring resonant wavelength to shift towards longer wavelengths. The thermal tuning unit features a large tuning range (nanometer-scale) and can be used as a coarse tuning unit. However, its response speed is limited by the thermal time constant of the microring (typically on the order of tens of microseconds), and it requires continuous power supply in steady state, resulting in high static power consumption.

[0036] The electrically tunable unit is a MOS capacitor-doped structure integrated on a microring resonant waveguide. It is constructed by forming heavily doped regions on both sides of the waveguide and sandwiching an intrinsic silicon layer in the middle. Its working principle is that when a forward bias voltage is applied to the capacitor structure, charge carriers accumulate on the waveguide surface, reducing the effective refractive index of the waveguide through plasma dispersion, thus shifting the microring resonant wavelength towards shorter wavelengths. The electrically tunable unit features fast response speed (GHz level), requires only a bias voltage to be maintained in steady state, and has extremely low static power consumption, making it suitable for fine-tuning. However, its single-device tuning range is limited (typically in the hundreds of picometers).

[0037] In this embodiment, the thermal tuning unit and the electrical tuning unit together constitute a vernier caliper-style hybrid tuning architecture with coarse and fine tuning. Utilizing their opposite wavelength offset directions and complementary tuning characteristics, it balances a wide tuning range with high-precision locking performance. Simultaneously, a low-bit digital-to-analog converter (DAC) can be used to drive both units separately, effectively achieving the tuning accuracy of a high-bit DAC and reducing system hardware costs. The resonant wavelength calibration method described in this embodiment uses ambient temperature as a criterion to dynamically adjust the allocation ratio of thermal and electrical tuning, optimizing system power consumption and operational reliability while ensuring wavelength locking accuracy. Specifically, it includes the following steps: Step S1: Periodically collect the ambient temperature of the micro-ring modulator at the first preset operating frequency.

[0038] In some embodiments, the microprocessor periodically reads the sampled values ​​of the thermistor via an analog-to-digital converter (ADC) to calculate the current ambient temperature. The sampling period is determined by a first preset operating frequency. In this step, the first preset operating frequency is lower than a second preset operating frequency, which is the operating frequency of the built-in temperature control module of the micro-ring modulator; that is, the sampling frequency of the upper-layer strategy calibration is lower than the operating frequency of the lower-layer built-in temperature control module. Preferably, the ambient temperature can be the current chip temperature of the micro-ring modulator or the external indoor temperature of the entire device.

[0039] The purpose is that the underlying PID temperature control is responsible for the rapid thermal tuning amplitude closed loop, which requires a high frequency to ensure the response speed, so that the thermal tuning unit can quickly achieve thermal tuning; while the upper-level calibration is responsible for the macroscopic switching of the tuning mode, which does not require an excessively high sampling frequency. A lower sampling frequency can reduce the computational load of the microprocessor and the power consumption of the system, while avoiding control coupling with the underlying high-frequency temperature control to prevent system oscillation.

[0040] In some embodiments, a sampling frequency adaptive adjustment step is further included, which dynamically adjusts the first preset operating frequency according to changes in ambient temperature, taking into account both system power consumption and response speed, specifically including: Step S11: Obtain the temperature change rate of the current cycle. The temperature change rate can be calculated by the temperature sampling values ​​of multiple consecutive cycles. For example, the difference between the temperature of the current cycle and the temperature of the previous cycle is divided by the sampling cycle duration to obtain the current temperature change rate.

[0041] Step S12: Compare the calculated temperature change rate with the first preset speed threshold and the second preset speed threshold respectively, and adjust the first preset operating frequency according to the comparison results. If the rate of temperature change is less than the first preset speed threshold, it indicates that the ambient temperature changes gradually with minimal fluctuations, and high-frequency sampling monitoring is unnecessary. In this case, the first preset operating frequency is reduced according to the second preset reduction. By reducing the sampling frequency, the computational overhead of the microprocessor and the power consumption of the ADC sampling are reduced, thus optimizing the overall energy efficiency of the system.

[0042] If the rate of temperature change is greater than or equal to the first preset rate threshold and less than the second preset rate threshold, it indicates that the temperature change is at a moderate rate, and the current sampling frequency is sufficient to match the rate of change. In this case, the current first preset operating frequency is maintained unchanged to ensure monitoring accuracy while avoiding system fluctuations caused by frequent frequency adjustments.

[0043] If the rate of temperature change is greater than or equal to the second preset rate threshold, it indicates a rapid change in ambient temperature, such as a sudden increase in GPU load causing a rapid rise in chip temperature in a CPO scenario. In this case, the first preset operating frequency is increased by the second preset increment to increase the temperature sampling density, so as to capture temperature changes more promptly, quickly adjust the tuning strategy, and avoid loss of resonance wavelength lock-up due to untimely response.

[0044] As mentioned earlier, the thermal tuning of the micro-ring modulator is also affected by the ambient temperature. Therefore, monitoring the ambient temperature is used as one of the indicators for thermal tuning control or tuning strategy switching of the micro-ring modulator. However, ambient temperature is dynamic and fluctuates differently under different operating conditions. If ambient temperature is considered throughout the tuning process, it will inevitably increase system power consumption and algorithm complexity or computational load. On the one hand, from the data acquisition perspective, the system frequently collects a large amount of temperature data from the temperature sensor. If the ambient temperature does not fluctuate significantly at the current stage, the frequently collected data will become a large amount of redundant data, thereby increasing system power consumption and occupying a large amount of system memory, wasting resources. If the ambient temperature fluctuates significantly at the current stage, the data collected by the traditional fixed sampling frequency is insufficient to reflect the current fluctuation, resulting in insufficient temperature accuracy even if ambient temperature is considered for thermal tuning. On the other hand, from the algorithm perspective, if the ambient temperature does not fluctuate significantly, the large amount of redundant data collected will greatly increase the computational load if no data processing is performed. If data processing is performed, such as data filtering, it will increase the algorithm complexity. Furthermore, if the thermal tuning amplitude is frequently adjusted or the tuning strategy is switched when the ambient temperature does not fluctuate significantly, it will also increase the system load and power consumption. Therefore, when incorporating ambient temperature into the thermal tuning process, a trade-off must be struck between the acquisition frequency (i.e., the first operating frequency) and data redundancy, as well as algorithm complexity and system power consumption. In other words, the system needs to balance tuning accuracy, response speed, and system power consumption.

[0045] In some other embodiments, the method further includes the step S13: after the frequency increase operation is performed, if the increased first preset working frequency does not exceed the threshold ratio of the second preset working frequency (preferably, in this embodiment the threshold ratio is set to 1 / 2), and the temperature change rate exceeds the third preset speed threshold, the sampling frequency is further increased and a temperature prediction mechanism is used to achieve an early response.

[0046] Of course, the sampling frequency cannot be increased indefinitely when the ambient temperature fluctuates greatly, as this would also increase the system load and generate a large amount of heat. Furthermore, if the ambient temperature changes too rapidly within a cycle, i.e., rises sharply, it actually indicates that the current computational workload may be huge or there may be an external emergency, such as an air conditioning failure. In this case, frequently collecting ambient temperature data and adjusting the thermal tuning amplitude accordingly is not very meaningful. For example, even if the system collects temperature data and adjusts the heating power at an extremely high frequency, it cannot keep up with the drift speed, and the control signal will lag significantly, resulting in the loss of tuning significance. Therefore, in this application, during the process of increasing the operating frequency, once the temperature change rate is detected to exceed the third preset speed threshold, the first preset operating frequency is no longer increased. Instead, it directly switches to pure electrical tuning, that is, ambient temperature is no longer included in the indicators for adjusting the thermal tuning process.

[0047] In other embodiments, the method further includes: Step S131: After pausing the increase in sampling frequency, the least squares method is used to perform linear regression fitting on the temperature sampling values ​​in the most recent M periods to calculate the current temperature change slope k and the fitting correlation coefficient r². Based on the comparison result between the fitting correlation coefficient and the preset confidence threshold, different judgment criteria are selected.

[0048] Where M is the preset fitting window length, which can be adjusted according to system characteristics; the fitting correlation coefficient r² is used to characterize the linearity of temperature change, and its value ranges from 0 to 1. The closer r² is to 1, the better the linearity of temperature change, and the higher the reliability of the prediction results based on linear fitting.

[0049] Step S132: If the fitted correlation coefficient r² is greater than the preset confidence threshold (in this embodiment, the preset confidence threshold is set to 0.95), it indicates that the current temperature exhibits a strict linear trend, and the prediction result has high reliability. At this time, the predicted temperature value for the next period can be calculated based on the current temperature change slope k, and this predicted temperature value can be used to replace the measured temperature to perform the aforementioned temperature range judgment step, thus switching the tuning scheme in advance.

[0050] The advantage of using a predictive mechanism is that it can overcome the limitations of the sampling period, predict the temperature change trend in advance, and complete the adjustment of the tuning strategy before the actual temperature reaches the threshold of the interval. This transforms the traditional post-feedback correction into pre-prediction compensation, effectively offsetting the thermal inertia lag of thermal tuning, significantly improving the system's response speed to rapid linear temperature changes, and avoiding instantaneous detuning.

[0051] In some embodiments, before calculating the predicted temperature value for the next cycle, an acceleration verification step is included to further identify the nonlinear characteristics of temperature changes and avoid prediction misjudgments when the temperature increases rapidly. Specifically, this includes: Step S1321: Calculate the temperature change acceleration of the current cycle. The temperature change acceleration can be calculated from the temperature change rate of multiple consecutive cycles.

[0052] For example, the difference between the current rate of temperature change and the rate of temperature change in the previous period, divided by the sampling period, yields the current acceleration of temperature change. A positive acceleration indicates that the rate of temperature change is accelerating, while a negative acceleration indicates that the rate of temperature change is slowing down.

[0053] Step S1322 determines whether to perform temperature prediction based on the comparison result between the acceleration value and the preset acceleration threshold.

[0054] Specifically, if the acceleration of temperature change is positive and greater than or equal to a preset acceleration threshold, it indicates that the temperature is not only changing rapidly, but the rate of change is also accelerating, exhibiting obvious nonlinear acceleration characteristics. In this case, the prediction result based on linear fitting will deviate significantly from the actual temperature, resulting in extremely low reliability. To avoid misjudgment, temperature prediction is not performed in this situation; instead, the temperature range judgment is performed directly based on the measured ambient temperature, ensuring the reliability of the calibration strategy.

[0055] If the acceleration due to temperature change is negative or less than the preset acceleration threshold, it indicates that the rate of temperature change is stabilizing or decelerating, the linearity of temperature change is good, and the error of linear prediction is within an acceptable range. At this point, the predicted temperature value for the next cycle is calculated based on the current temperature change slope k, and this predicted temperature value is used to determine the temperature range.

[0056] For example, when the cooling fan of the CPO system suddenly starts, the ambient temperature will change from a rapid rise to a decelerating rise. At this time, the acceleration is negative, and the temperature change gradually tends to be stable, so linear prediction has high reliability. However, when the GPU load continues to increase step by step, the acceleration of temperature rise is continuously positive and large. At this time, the temperature change is highly nonlinear, and prediction mechanism is not suitable.

[0057] In step S133, if the fitted correlation coefficient r² is less than or equal to the preset confidence threshold, it indicates that the current temperature fluctuations are chaotic, the nonlinear characteristics are obvious, the linear fitting error is large, and the prediction result is unreliable. At this time, temperature prediction is abandoned, and the temperature range judgment is directly performed based on the current measured ambient temperature to ensure the accuracy of the calibration strategy and avoid erroneous switching caused by incorrect prediction.

[0058] Furthermore, the reason for setting a frequency upper limit (half of the second preset operating frequency) is that if the upper-level calibration frequency is too high, approaching or even exceeding half of the lower-level temperature control frequency, it will cause a beat frequency effect with the lower-level PID closed loop, leading to control oscillation and reducing system stability. Therefore, by setting a frequency upper limit, the upper-level calibration frequency is always kept far below the lower-level temperature control frequency, maintaining the decoupling state of the two control layers.

[0059] Step S2: Perform temperature range judgment, compare the currently collected ambient temperature with the pre-stored first preset temperature threshold T1 and second preset temperature threshold T2 respectively, and determine the range to which the current temperature belongs.

[0060] In this embodiment, the first preset temperature threshold T1 is the boundary between the low temperature range and the transition range, and the second preset temperature threshold T2 is the boundary between the transition range and the high temperature range, with T1 < T2. The two thresholds can be pre-calibrated and stored based on the rated operating temperature of the micro-ring modulator, the maximum tuning range of the electronic tuning unit, and the heat dissipation conditions of the CPO system.

[0061] Step S3: Execute the corresponding calibration strategy according to the different temperature ranges: Step S31, First case: If the current ambient temperature is less than the first preset temperature threshold T1, it is determined that the system is in the low temperature operating range, and a maintenance operation is performed to maintain the current tuning scheme of thermal tuning and electrical tuning, and keep the target thermal tuning amplitude and the target electrical tuning amplitude unchanged.

[0062] In this range, the ambient temperature is low, the overall heat load of the micro-ring chip is small, and the additional heating of the thermal tuning unit will not cause overheating risk. Moreover, thermal tuning has a wide tuning range, and thermoelectric synergy can cover the entire range of wavelength deviations while ensuring sufficient tuning margin. Maintaining the original synergy scheme at this time can fully utilize the range advantage of thermal tuning, while electrical tuning retains sufficient fine-tuning margin to cope with small temperature fluctuations.

[0063] Step S32, the second case: if the current ambient temperature is greater than or equal to the first preset temperature threshold T1 and less than the second preset temperature threshold T2, the system is determined to be in the mid-temperature transition range. A gradual adjustment operation of the tuning amplitude is performed. The target thermal tuning amplitude of the thermal tuning unit is reduced according to the preset first reduction, and the target electrical tuning amplitude of the electrical tuning unit is increased accordingly based on the reduced target thermal tuning amplitude, so as to ensure that the total tuning amount remains unchanged, that is, the total tuning amount always matches the deviation between the resonant wavelength and the target operating wavelength. Then, tuning is performed according to the updated target thermal tuning amplitude and target electrical tuning amplitude.

[0064] In this range, the ambient temperature has already risen to a certain level. If a high thermal tuning power is maintained, it will exacerbate local heat accumulation on the chip, which will not only increase the overall power consumption of the system but may also cause thermal crosstalk between adjacent channels, and even affect the reliability of the device. Therefore, by gradually reducing the proportion of thermal tuning and increasing the proportion of electrical tuning, without changing the total tuning amount and ensuring wavelength lock stability, heating power consumption is reduced, the thermal load is lowered, and a smooth transition from full thermoelectric synergy to pure electrical tuning is achieved.

[0065] More specifically, the preset first reduction utilizes the following formula: ΔA = A1 × (T2 - Tc) / (T2 - T1) is calculated; where ΔA is the preset first reduction, that is, the thermal tuning amplitude that needs to be reduced in the current cycle; A1 is the initial target thermal tuning amplitude, that is, the reference thermal tuning amplitude in the low temperature range; and Tc is the current measured ambient temperature.

[0066] The reduction calculated using this linear formula allows for a linear decrease in the thermal tuning amplitude as temperature increases. At temperature T1, the reduction is zero, and the thermal tuning amplitude remains at its initial value. When the temperature rises to T2, the reduction equals the initial thermal tuning amplitude, and the thermal tuning amplitude drops to zero. At intermediate temperatures, the reduction is smooth and proportional. Correspondingly, the electrical tuning amplitude increases linearly with temperature, while the total tuning amplitude remains constant, thus preventing abrupt changes in the resonant wavelength during tuning mode switching and ensuring the quality of optical signal transmission.

[0067] Step S33, the third case: if the current ambient temperature is greater than or equal to the second preset temperature threshold T2, it is determined that the system is in the high temperature operating range, and a mode switching operation is performed to switch the current tuning scheme to a pure electric tuning scheme, relying entirely on the electric tuning unit to maintain wavelength lock.

[0068] Furthermore, in some other embodiments, in step S33, during the switching process of switching the current tuning scheme to the pure electric tuning scheme, a soft switching mechanism is adopted instead of directly shutting off the thermal tuning and jumping the electric tuning parameters. For example, the thermal tuning power is linearly or exponentially decayed to zero with a preset decreasing slope, while the electric tuning bias is synchronously increased with a preset increasing slope. The two changes are strictly synchronized to ensure that the total tuning amount remains constant throughout the switching process.

[0069] The purpose of setting up soft switching is that the micro-ring thermal tuning has thermal inertia, and its temperature change lags behind the change in driving power. If the power changes abruptly, it will cause the resonant wavelength to jump instantaneously, causing signal errors. The synchronous and gradual soft switching method can match the thermal response characteristics of the micro-ring, realize wavelength switching without impact, and ensure the continuity and stability of the communication link.

[0070] In some embodiments, the total time for the thermally tuned power to decay from its current value to zero is greater than 0.5 times and less than 3 times the thermal time constant of the micro-ring modulator. If the time is less than 0.5 times the thermal time constant, the switching speed is too fast, and thermal inertia will cause the actual temperature to lag behind the power change, resulting in wavelength fluctuations. If the time is greater than 3 times the thermal time constant, the switching process is too slow, failing to respond promptly to high-temperature conditions and prolonging the system's transition time, which is detrimental to rapid stabilization. Controlling the time within the range of 0.5 to 3 times the thermal time constant balances switching smoothness and response speed.

[0071] When the system switches to a pure electric tuning scheme, a lockout fault protection mechanism is also set up, which includes: The system monitors the target tuning amplitude of the electronic tuning unit in real time. If the target tuning amplitude reaches the preset safety limit and the duration exceeds the preset time threshold, and the resonant wavelength still cannot converge to the target operating wavelength, the system is determined to have lost lock fault.

[0072] After the fault protection is triggered, the current wavelength locking process is first paused, the thermal tuning power is forced to zero, and heating is stopped. Then, wait for the preset thermal equilibrium time (in this embodiment, the preset thermal equilibrium time is set to 1ms, which matches the thermal time constant of the micro-ring). After the temperature of the micro-ring waveguide has completely dropped to the ambient temperature and the thermal field has reached a stable equilibrium state, all the steps of this calibration method are executed again from the beginning to perform a new round of wavelength search and calibration.

[0073] This fault protection mechanism can handle the problem of loss of lock under extreme operating conditions, such as abnormally high ambient temperature exceeding the maximum range of electrical tuning, or deviation of the tuning range caused by device process deviation. It avoids the system from falling into a dead loop of repeated adjustments but unable to lock. At the same time, it resets the device by thermal adjustment to zero, protecting the device from damage caused by long-term overvoltage drive, and improving the fault tolerance and long-term reliability of the system.

[0074] Example 2: This calibration method can be applied to multi-wavelength optical emission systems containing N parallel micro-ring modulators, such as wavelength division multiplexing silicon photonic emission arrays. For multi-channel scenarios, this method employs a time-division polling calibration mechanism to suppress thermal crosstalk between channels, specifically including: During a complete calibration cycle, only one micro-loop modulator undergoes an update operation on its tuning amplitude, while the remaining N-1 micro-loop modulators maintain their current thermal tuning power and remain in a frozen state.

[0075] After completing the calibration update of the current channel, the system will switch to the next micro-ring modulator at a preset polling interval to perform calibration, until all N micro-rings have completed one calibration, thus completing a full calibration cycle.

[0076] In this embodiment, the multi-wavelength optical emission system includes at least one optical propagation path, and each optical propagation path has at least one micro-ring modulator arranged in parallel along its optical transmission direction, with each micro-ring modulator having a different target operating wavelength. Preferably, when the multi-wavelength optical emission system includes N optical propagation paths (N≥2), for example, when one optical input signal is divided into N optical propagation paths by an optical beam splitter, the multi-wavelength optical emission system includes N parallel micro-ring modulators.

[0077] The preset polling interval is greater than three times the thermal time constant of the micro-ring modulator. This is because after a single micro-ring adjusts its thermal tuning power, the heat it generates diffuses to the surrounding area, causing thermal crosstalk to adjacent channels. It takes time for the thermal field to stabilize again; a thermal time constant of three times is sufficient to completely dissipate the temperature fluctuations caused by the previous channel's heating adjustment, ensuring that the surrounding thermal field is stable when calibrating the next channel, thus avoiding thermal interference from adjacent channels affecting calibration accuracy.

[0078] This time-division polling mechanism can effectively solve the problem of inter-channel thermal crosstalk in high-density micro-ring arrays from the control level, avoid problems such as mutual thermal field interference and wavelength-locked oscillation caused by simultaneous tuning of multiple channels, significantly improve the overall stability of multi-channel parallel systems, and adapt to the application requirements of high-density integration of hundreds of optical channels in CPO scenarios.

[0079] Example 3: like Figure 2 As shown, the present invention also provides a resonant wavelength calibration system for a micro-ring modulator, comprising: An ambient temperature acquisition module is used to periodically acquire the ambient temperature of the micro-ring modulator at a first preset operating frequency. The temperature range determination module is used to compare the current ambient temperature with the first preset temperature threshold T1 and the second preset temperature threshold T2 respectively to determine the range to which the current temperature belongs. The tuning scheme switching module is used to maintain the current tuning scheme that combines thermal tuning and electrical tuning when the current ambient temperature is lower than the first preset temperature threshold T1. When the current ambient temperature is greater than or equal to the first preset temperature threshold T1 and less than the second preset temperature threshold T2, the target thermal tuning amplitude of the thermal tuning unit is reduced according to the preset first reduction, and the target electrical tuning amplitude of the electrical tuning unit is increased according to the reduced target thermal tuning amplitude; tuning is performed according to the reduced target thermal tuning amplitude and the increased target electrical tuning amplitude. If the current ambient temperature is greater than or equal to the second preset temperature threshold T2, the current tuning scheme will be switched to the pure electric tuning scheme.

[0080] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0081] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a computer terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0082] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A method for calibrating the resonant wavelength of a micro-ring modulator, characterized in that, The micro-ring modulator integrates a thermal tuning unit and an electrical tuning unit, which work together to compensate for the deviation between the resonant wavelength and the target operating wavelength. The calibration method includes the following steps: The ambient temperature of the micro-ring modulator is periodically collected at a first preset operating frequency. The current ambient temperature is compared with the first preset temperature threshold T1 and the second preset temperature threshold T2 to determine the range to which the current ambient temperature belongs. If the current ambient temperature is less than the first preset temperature threshold T1, the current tuning scheme of thermal tuning and electrical tuning will be maintained. If the current ambient temperature is greater than or equal to the first preset temperature threshold T1 and less than the second preset temperature threshold T2, then the target thermal tuning amplitude of the thermal tuning unit is reduced according to the preset first reduction, and the target electrical tuning amplitude of the electrical tuning unit is increased according to the reduced target thermal tuning amplitude; and coordinated tuning is performed according to the reduced target thermal tuning amplitude and the increased target electrical tuning amplitude. If the current ambient temperature is greater than or equal to the second preset temperature threshold T2, the current tuning scheme will be switched to the pure electric tuning scheme.

2. The resonant wavelength calibration method for a micro-ring modulator according to claim 1, characterized in that, The first preset operating frequency is less than the second preset operating frequency, which is the operating frequency of the built-in temperature control module of the micro-ring modulator.

3. The resonant wavelength calibration method for a micro-ring modulator according to claim 2, characterized in that, It also includes a sampling frequency adaptive adjustment step: Obtain the temperature change rate of the current cycle and compare the temperature change rate with the first preset speed threshold and the second preset speed threshold. If the rate of temperature change is less than the first preset rate threshold, the first preset operating frequency is reduced according to the second preset reduction. If the rate of temperature change is greater than or equal to the first preset speed threshold and less than the second preset speed threshold, the current first preset operating frequency is maintained unchanged. If the rate of temperature change is greater than or equal to the second preset speed threshold, the first preset operating frequency is increased by the second preset increment.

4. The resonant wavelength calibration method for a micro-ring modulator according to claim 3, characterized in that: When the increased first preset working frequency does not exceed the threshold ratio of the second preset working frequency, and the temperature change rate exceeds the third preset speed threshold, the increase of the sampling frequency is paused, and the current temperature change slope k and the fitting correlation coefficient r² are calculated by linear regression of the temperature values ​​within M cycles. If the fitted correlation coefficient r² is greater than the preset confidence threshold, the predicted temperature value for the next cycle is calculated based on the current temperature change slope k, and the temperature range is judged based on the predicted temperature value to switch the tuning scheme. If the fitted correlation coefficient r² is less than or equal to the preset threshold, the temperature range is determined directly based on the current measured ambient temperature to switch the tuning scheme.

5. The resonant wavelength calibration method for a micro-ring modulator according to claim 4, characterized in that, Before calculating the predicted temperature value for the next cycle, an acceleration verification step is also included: Calculate the acceleration of temperature change in the current cycle; If the acceleration due to temperature change is positive and greater than or equal to the preset acceleration threshold, then the temperature range judgment is performed based on the measured ambient temperature. If the acceleration due to temperature change is negative or less than the preset acceleration threshold, the predicted temperature value for the next cycle is calculated based on the current temperature change slope k, and the temperature range is determined using this predicted temperature value.

6. The resonant wavelength calibration method for a micro-ring modulator according to claim 1, characterized in that, The total time it takes for the thermally tuned power to decay from its current value to zero is greater than 0.5 times the thermal time constant of the micro-ring modulator, but less than 3 times the thermal time constant of the micro-ring modulator.

7. The resonant wavelength calibration method for a micro-ring modulator according to claim 1, characterized in that, The preset first reduction and utilization formula: ΔA = A1 × (T2 - Tc) / (T2 - T1) is calculated; Where ΔA is the preset first reduction, A1 is the initial target thermal tuning amplitude, and Tc is the current ambient temperature.

8. The resonant wavelength calibration method for a micro-ring modulator according to claim 1, characterized in that, After switching to the pure electric tuning scheme, if the target electric tuning amplitude of the electric tuning unit reaches the preset safety limit and the resonant wavelength still cannot converge after the duration exceeds the preset time threshold, then wavelength locking is paused, the thermal tuning power is reduced to zero, and after the preset thermal equilibrium time, recalibration is performed.

9. The resonant wavelength calibration method for a micro-ring modulator according to claim 1, characterized in that, When it is determined that the current ambient temperature is greater than or equal to the second preset temperature threshold T2, and during the switching process from the current tuning scheme to the pure electric tuning scheme, the thermal tuning power is linearly or exponentially decayed to zero with a preset decreasing slope, while the electric tuning bias is synchronously increased with a preset increasing slope.

10. A resonant wavelength calibration system for a micro-ring modulator, characterized in that, include: An ambient temperature acquisition module is used to periodically acquire the ambient temperature of the micro-ring modulator at a first preset operating frequency. The temperature range determination module is used to compare the current ambient temperature with the first preset temperature threshold T1 and the second preset temperature threshold T2 respectively to determine the range to which the current temperature belongs. The tuning scheme switching module is used to maintain the current tuning scheme that combines thermal tuning and electrical tuning when the current ambient temperature is lower than the first preset temperature threshold T1. Furthermore, when the current ambient temperature is greater than or equal to the first preset temperature threshold T1 and less than the second preset temperature threshold T2, the target thermal tuning amplitude of the thermal tuning unit is reduced according to the preset first reduction, and the target electrical tuning amplitude of the electrical tuning unit is increased according to the reduced target thermal tuning amplitude. Tuning is performed according to the reduced target thermal tuning amplitude and the increased target electrical tuning amplitude. And when the current ambient temperature is greater than or equal to the second preset temperature threshold T2, the current tuning scheme will be switched to the pure electric tuning scheme.

Images