Signal calibration methods, apparatuses, media, and program products

Abstract

The application relates to the chip technical field and provides a signal calibration method, equipment, medium and program product.The method comprises the following steps: when a data signal is received in a main receiving channel, the same data signal is received by using a redundant receiving channel which is parallel to the main receiving channel, so that at least one set of effective parameter combinations capable of correctly receiving the data signal is monitored and acquired by the redundant receiving channel; an optimal parameter combination is determined from the effective parameter combinations according to the positions of the effective parameter combinations in an effective parameter window; and the optimal parameter combination is applied to the main receiving channel to calibrate the data signal of the main receiving channel.In the application, the monitoring and calibration processes are independently carried out on the redundant channel, and the service data transmission of the main receiving channel is not affected, so that the problem that the traditional static parameter configuration cannot adapt to the dynamic changes of temperature, voltage and other environments and causes the signal quality to decrease is solved, and the system performance loss and the data loss risk caused by the interruption of the service for parameter adjustment are avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the chip technical field, and particularly to a signal calibration method, device, medium and program product. BACKGROUND

[0002] High-speed signal quality monitoring and calibration is a key technology to guarantee the reliability and performance of signal transmission in complex electronic systems.

[0003] At present, external instruments such as high-bandwidth oscilloscopes are usually used to directly measure signal waveforms, and parameters such as eye diagrams and bit error rates are analyzed to evaluate signal quality, but this method requires expensive external instruments, resulting in high costs. In addition, this method is difficult to implement online monitoring, has poor flexibility, and the test process is time-consuming, which reduces the debugging efficiency and is not suitable for mass production testing. SUMMARY

[0004] The present application provides a signal calibration method, device, medium and program product to solve the defects in the related art.

[0005] The present application provides a signal calibration method, comprising the following steps:

[0006] When receiving a data signal in a main receiving channel, a redundant receiving channel parallel to the main receiving channel is used to receive the same data signal, so that the redundant receiving channel monitors and obtains at least one set of valid parameter combinations that can correctly receive the data signal;

[0007] According to the position of each valid parameter combination in the valid parameter window, the optimal parameter combination is determined from each valid parameter combination; the valid parameter window refers to the parameter region capable of correctly receiving the data signal formed by all valid parameter combinations;

[0008] The optimal parameter combination is applied to the main receiving channel to calibrate the data signal of the main receiving channel.

[0009] According to the signal calibration method provided by the present application, the determination of the valid parameter combination comprises:

[0010] Each parameter combination composed of a reference voltage and a data sampling clock phase is traversed, and the data signal reception under each parameter combination is tested;

[0011] The parameter combination capable of correctly receiving the data signal is determined as the valid parameter combination.

[0012] According to the signal calibration method provided by the present application, the traversal of each parameter combination composed of a reference voltage and a data sampling clock phase, and the testing of the data signal reception under each parameter combination, comprises:

[0013] The optimal parameter combination obtained by the circuit training or the last determined optimal parameter combination is used as a starting point parameter combination;

[0014] The reference voltage in the starting point parameter combination is fixed, the data sampling clock phase is scanned to the left and to the right respectively, the data sampling clock phase range capable of correctly receiving the data signal under the fixed reference voltage is obtained, and the fixed reference voltage and each data sampling clock phase in the data sampling clock phase range are constructed into an effective parameter combination under this scanning;

[0015] The fixed reference voltage is changed, and the step of scanning the data sampling clock phase is returned to be executed until the scanned data sampling clock phase range under the fixed reference voltage is zero, and all effective parameter combinations are obtained.

[0016] According to the signal calibration method provided by the application, the optimal parameter combination is determined from each effective parameter combination according to the position of each effective parameter combination in the effective parameter window, which comprises:

[0017] The geometric center of the effective parameter window in the reference voltage dimension and the data sampling clock phase dimension is calculated;

[0018] The effective parameter combination corresponding to the geometric center is used as the optimal parameter combination.

[0019] According to the signal calibration method provided by the application, the geometric center of the effective parameter window in the reference voltage dimension and the data sampling clock phase dimension is calculated, which comprises:

[0020] The maximum reference voltage and the minimum reference voltage in all effective parameter combinations are determined, the average value of the maximum reference voltage and the minimum reference voltage is used as an optimal reference voltage, and the optimal reference voltage is used as the geometric center in the reference voltage dimension;

[0021] The maximum data sampling clock phase and the minimum data sampling clock phase in all effective parameter combinations are determined, the average value of the maximum data sampling clock phase and the minimum data sampling clock phase is used as an optimal data sampling clock phase, and the optimal data sampling clock phase is used as the geometric center in the data sampling clock phase dimension.

[0022] According to the signal calibration method provided by the application, the optimal parameter combination is applied to the main receiving channel, which comprises:

[0023] After the reference voltage of the main receiving channel is adjusted to the optimal reference voltage value in the optimal parameter combination, the data sampling clock phase of the main receiving channel is adjusted to the optimal data sampling clock phase in the optimal parameter combination;

[0024] or,

[0025] adjusting the reference voltage of the main receiving channel to the optimal reference voltage value in the optimal parameter combination after adjusting the data sampling clock phase of the main receiving channel to the optimal data sampling clock phase in the optimal parameter combination;

[0026] or,

[0027] adjusting the current parameter combination of the main receiving channel to the optimal parameter combination once.

[0028] According to the signal calibration method provided by the application, the application of the optimal parameter combination to the main receiving channel comprises:

[0029] switching the data of the main receiving channel to the redundant receiving channel for transmission;

[0030] switching the data of the redundant receiving channel back to the main receiving channel after the adjustment of the current parameter combination of the main receiving channel to the optimal parameter combination.

[0031] The application further provides a signal calibration device, comprising the following modules:

[0032] a monitoring unit, configured to receive the same data signal by a redundant receiving channel in parallel with the main receiving channel when the main receiving channel receives the data signal, so that the redundant receiving channel monitors at least one valid parameter combination capable of correctly receiving the data signal;

[0033] a determining unit, configured to determine an optimal parameter combination from the valid parameter combinations according to the positions of the valid parameter combinations in a valid parameter window; the valid parameter window refers to a parameter region capable of correctly receiving the data signal formed by all the valid parameter combinations;

[0034] a calibration unit, configured to apply the optimal parameter combination to the main receiving channel to calibrate the data signal of the main receiving channel.

[0035] The application further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the signal calibration method according to any one of the above when executing the program.

[0036] The application further provides a non-transitory computer readable storage medium, which stores a computer program executable by a processor to implement the signal calibration method according to any one of the above.

[0037] The application further provides a computer program product, comprising a computer program executable by a processor to implement the signal calibration method according to any one of the above.

[0038] The signal calibration method, device, medium and program product provided by the application, by setting a redundant receiving channel parallel to the main receiving channel, without interrupting the data transmission of the main receiving channel, the optimal parameter combination in the current link state is calculated by scanning and monitoring using the redundant receiving channel, and the optimal parameter combination is applied to the main receiving channel, realizing real-time and dynamic monitoring and calibration of the high-speed signal link. Since the monitoring and calibration process is carried out independently on the redundant channel, the business data transmission of the main receiving channel is not affected, thereby solving the problem that the traditional static parameter configuration cannot adapt to the dynamic changes of temperature, voltage and other environments, resulting in the decline of signal quality, and avoiding the system performance loss and data loss risk caused by the interruption of business for parameter adjustment, greatly improving the reliability and robustness of the high-speed signal transmission system. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the application or the related art, the following will briefly introduce the drawings needed to be used in the embodiments or the related art description. Obviously, the drawings in the following description are some embodiments of the application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0040] Figure 1 is a flowchart of the signal calibration method provided by the application.

[0041] Figure 2 is a UCIe receiving end circuit topology diagram provided by the application.

[0042] Figure 3 is a schematic diagram of all parameter combinations to be tested provided by the application.

[0043] Figure 4 is a parameter adjustment diagram provided by the application.

[0044] Figure 5 is a schematic diagram of the switching circuit of the main receiving channel and the redundant receiving channel provided by the application.

[0045] Figure 6 is a structural diagram of the signal calibration device provided by the application.

[0046] Figure 7 is a structural diagram of the electronic device provided by the application. DETAILED DESCRIPTION

[0047] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below in conjunction with the drawings in the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0048] In a high-speed digital communication system, it is crucial to ensure the reliability and performance of signal transmission. As the data rate continues to increase to the Gb / s level, signals are more susceptible to channel loss, temperature drift, and noise interference, resulting in a decline in signal quality. Therefore, real-time and accurate quality monitoring of high-speed signals has become a key technology.

[0049] In related technologies, a common signal calibration method is to directly measure the signal through external instruments such as high-bandwidth oscilloscopes. This method inputs the measured signal into the oscilloscope, forms an eye diagram by superimposing waveforms of millions of bit periods, and analyzes parameters such as eye height and eye width in combination with bit error rate test results to evaluate signal quality. However, this method has significant drawbacks: first, high-bandwidth oscilloscopes and supporting probe systems are expensive, resulting in high testing costs; second, this method relies on external devices and cannot be integrated into the chip, making it difficult to achieve online monitoring and poor flexibility; in addition, generating an eye diagram requires collecting a large amount of data, which is time-consuming and inefficient for debugging, making it unsuitable for mass production testing scenarios.

[0050] To this end, the present application provides a signal calibration method aimed at solving the problem of relying on expensive external devices, being unable to monitor online, and having low testing efficiency in related technologies, thereby enabling low-cost, integrable, and efficient online high-speed signal calibration.

[0051] Figure 1 The present application provides a flowchart of the signal calibration method, as shown in Figure 1 The method comprises steps 110, 120 and 130.

[0052] Step 110, when receiving a data signal in a main receiving channel, the same data signal is received by a redundant receiving channel parallel to the main receiving channel, so that the redundant receiving channel monitors and obtains at least one set of valid parameter combinations capable of correctly receiving the data signal.

[0053] Here, the main receiving channel can be understood as a physical channel responsible for actual business data transmission in a high-speed interface. Among them, Figure 2 The present application provides a UCIe receiving end circuit topology diagram, as shown in Figure 2As shown in the UCIe receiver (RX) circuit topology, the main receiving channels can be one or more lanes, such as lane0 to lane63, which continuously receive data packets (packet0 to packet63) from the transmitter (TX).

[0054] Correspondingly, the redundant receiving channel can be understood as a backup channel that is arranged in parallel with the main receiving channel and has the same or similar hardware structure and function as the main receiving channel. For example, Figure 2 As shown, each main receiving channel lane corresponds to a redundant receiving channel lane_rd, for example, lane0 corresponds to lane0_rd. In normal operation, the redundant receiving channel does not carry critical service data, but is used to perform monitoring, testing or calibration and other auxiliary functions. In this embodiment, it simulates the working state of the main receiving channel by receiving the same data signal as the main receiving channel, and optimizes the parameters on this basis.

[0055] The data signal can refer to any information carried on the high-speed link, which can be a pseudo-random binary sequence used for link training during system power-on or link initialization, or actual service data transmitted after the link enters the ACTIVE state, which is not limited in this embodiment. Among them,

[0056] The effective parameter combination can be understood as a set of receiver parameter settings that can enable the receiver to correctly decode the data signal. In high-speed signal reception, the key parameters usually include reference voltage (Vref) and data sampling clock phase (deskew). The reference voltage is used to set the decision threshold of the decision device for signal levels "0" and "1", while the data sampling clock phase is used to determine the optimal sampling time within a bit period. Under the Vref and deskew in the effective parameter combination, the error rate of the data signal at the receiver is lower than a certain preset threshold.

[0057] By performing the above scanning monitoring on the redundant receiving channel, the main receiving channel can continue to transmit service data without interruption, ensuring the continuity of system operation. In other words, the signal calibration process of the present embodiment can be performed in real time when the link transmits actual service data, without interrupting data transmission.

[0058] Step 120, determining the optimal parameter combination from the effective parameter combinations according to the position of each effective parameter combination in the effective parameter window; the effective parameter window refers to the parameter region formed by all effective parameter combinations that can correctly receive the data signal.

[0059] Specifically, it is considered that although the valid parameter combination can ensure correct reception of the data signal at the current moment, part of the valid parameter combinations can be located at the edge position of the valid parameter window, but its anti-interference ability is weak, and a little noise or jitter can make it invalid parameter, thereby causing transmission error. The valid parameter window can be understood as a geometric region formed by all valid parameter combinations scanned and obtained in step 110 in a two-dimensional parameter coordinate system.

[0060] Generally, the closer the position of the valid parameter combination in the valid parameter window to the boundary of the valid parameter window, the smaller the working margin of the corresponding valid parameter combination, at this time, even a little external interference (such as voltage noise or clock jitter) can easily cause it to deviate from the valid parameter window, causing data reception error. The farther the position of the valid parameter combination in the valid parameter window to the boundary of the valid parameter window, the larger the working margin of the corresponding valid parameter combination, at this time, it can withstand a larger range of external interference, still can stably keep in the valid parameter window, ensure the reliability of data transmission.

[0061] Based on this, the embodiment determines the optimal parameter combination from each valid parameter combination according to the position of each valid parameter combination in the valid parameter window. The optimal parameter combination can be understood as the parameter combination that is the most robust and has the strongest interference ability among all valid parameter combinations. Preferably, the valid parameter combination corresponding to the geometric center of the valid parameter window generally has the largest and most balanced working margin in each parameter dimension, that is, the parameter combination is farthest from the window boundary as the invalid boundary in the whole, can resist the comprehensive interference from the voltage, clock and other dimensions to the greatest extent, and then the valid parameter combination corresponding to the geometric center of the valid parameter window can be taken as the optimal parameter combination.

[0062] As an optional embodiment, the minimum value vref_min and the maximum value vref_max of the reference voltage can be found in all valid parameter combinations, and then the average value vref_best=(vref_min+vref_max) / 2 is taken. Then, in all valid parameter combinations, the minimum value deskew_min and the maximum value deskew_max of the data sampling clock phase under the specific reference voltage vref_best are found, and then the average value deskew_best=(deskew_min+deskew_max) / 2 is taken. Finally, the combination (vref_best, deskew_best) is the optimal parameter combination determined this time.

[0063] As another optional embodiment, the widest deskew range can be determined among all the Vref in all the valid parameter combinations, and the midpoint of the deskew range is determined as deskew_best; then the center point of the corresponding Vref range is determined as vref_best at the specific data sampling clock phase of deskew_best. Finally, the combination (vref_best, deskew_best) is the optimal parameter combination determined this time.

[0064] Step 130, apply the optimal parameter combination to the main receiving channel to calibrate the data signal of the main receiving channel.

[0065] Specifically, since the optimal parameter combination is the most robust and the most interference-capable parameter combination among all the valid parameter combinations, after the optimal parameter combination is applied to the main receiving channel, the main receiving channel can be maximally resistant to signal characteristic deterioration caused by factors such as temperature, voltage, aging and noise, and the robustness and reliability of signal reception of the main receiving channel can be significantly improved, and the bit error rate of long-term operation can be reduced.

[0066] The application of the optimal parameter combination to the main receiving channel can be understood as updating the setting value of the hardware register or circuit inside the main receiving channel for controlling the reference voltage and the data sampling clock phase to the optimal parameter combination determined in step 120.

[0067] As an optional embodiment, the reference voltage Vref of the main receiving channel can be gradually adjusted from the current value to the optimal reference voltage vref_best in the optimal parameter combination, and after stabilization, the data sampling clock phase deskew is gradually adjusted from the current value to the optimal data sampling clock phase deskew_best in the optimal parameter combination. Alternatively, deskew can be adjusted first, and then Vref can be adjusted. The above-mentioned manner reduces the possibility of introducing transient errors by too large single adjustment amplitude through smooth parameter transition.

[0068] As another optional embodiment, the parameters of the main receiving channel can be directly updated from the current value to the optimal parameter combination (vref_best, deskew_best) at one time, and this way is the fastest in adjustment speed, but it can cause a large disturbance to signal decision at the moment of parameter switching.

[0069] As a preferred embodiment, the data stream of the main receiving channel can be temporarily switched to the redundant receiving channel (rd_lane) for transmission through a multiplexer (MUX). Since the parameters of the redundant receiving channel have been set to a known effective value (which can be the optimal value found this time), data transmission can continue seamlessly. Then, when the main receiving channel (lane) no longer carries real-time data stream, the parameter combination is adjusted to the optimal parameter combination. After the adjustment is completed, the data stream is switched back to the calibrated main receiving channel through the MUX. This "switching, adjusting, and switching back" method can achieve zero interference to the business data and realize truly uninterrupted calibration.

[0070] In actual applications, the specific application method can be selected by register configuration to adapt to different system requirements for adjustment speed and stability.

[0071] The signal calibration method provided by the embodiment can scan, monitor, and calculate the optimal parameter combination under the current link state by setting a redundant receiving channel parallel to the main receiving channel without interrupting the data transmission of the main receiving channel, and apply the optimal parameter combination to the main receiving channel, thereby realizing real-time and dynamic monitoring and calibration of the high-speed signal link. Since the monitoring and calibration process is carried out independently on the redundant channel, the business data transmission of the main receiving channel is not affected, thereby solving the problem that the traditional static parameter configuration cannot adapt to the dynamic changes of temperature, voltage, and other environments, leading to a decline in signal quality, and avoiding the system performance loss and data loss risk caused by interrupting the business for parameter adjustment, greatly improving the reliability and robustness of the high-speed signal transmission system.

[0072] Based on the above embodiment, the determination of the effective parameter combination includes:

[0073] Each parameter combination is tested for data signal reception under each parameter combination by traversing each parameter combination composed of the reference voltage and the data sampling clock phase.

[0074] The parameter combination that can correctly receive the data signal is determined as the effective parameter combination.

[0075] Specifically, the traversal can be understood as a systematic and regular scanning of the two-dimensional parameter space composed of the reference voltage (Vref) and the data sampling clock phase (deskew) in the preset range, and the scanning process aims to test every discrete parameter point in the space without omission.

[0076] As an optional embodiment, the scanning range and step can be determined first. For example, based on the design specification of the circuit and historical experience data, the scanning range of the reference voltage is determined as [Vref_start, Vref_end] and the step is ΔV, the scanning range of the data sampling clock phase is determined as [deskew_start, deskew_end] and the step is Δd. Then, one parameter is fixed and the other parameter is scanned. For example, the reference voltage is set as the starting value Vref_start, and then the scanning range of the data sampling clock phase is traversed with the step Δd, and each (Vref_start, deskew) parameter combination is tested. After the traversal with the starting value Vref_start is completed, the reference voltage is increased by one step, i.e., Vref_start+ΔV is used as the fixed reference voltage, and the scanning range of the data sampling clock phase is traversed again. Such a cycle is repeated until the reference voltage reaches Vref_end, so as to complete the traversal of the entire two-dimensional parameter space and obtain multiple parameter combinations.

[0077] Then, the data signal receiving condition under each parameter combination is tested, i.e., it is judged whether the data signal can be correctly received when the receiver of the redundant receiving channel is set as the current parameter combination. If yes, the corresponding parameter combination is used as the effective parameter combination.

[0078] As a specific embodiment, the above test can be completed by the bit error rate test circuit built in the redundant receiving channel, which can compare the received data with the known expected data sequence in real time, count the number of different bits, and thus calculate the bit error rate under the current parameter combination.

[0079] Among them, whether the data signal can be correctly received can be understood as whether the input data signal can be decoded without error, such as the bit error rate being lower than a preset threshold.

[0080] Figure 3 is the schematic diagram of all tested parameter combinations provided by the present application, as shown in Figure 3 The point marked as "O" is the point that passes the test and is determined as the effective parameter combination, and the point marked as "X" represents the invalid parameter combination that fails the test and has an excessive bit error rate. After the traversal scanning is completed, all the recorded "O" points collectively constitute the effective parameter combination in step 110.

[0081] By the above-mentioned way of traversing scanning and error code testing on the two-dimensional parameter space of the reference voltage and the data sampling clock phase, the complete and effective parameter window in the current link state can be accurately depicted, which not only ensures that the found parameter combination is indeed effective, but also provides a comprehensive and quantitative data basis for subsequent calculation of the optimal parameter combination, avoids suboptimal selection caused by local search or heuristic algorithm, and further improves the accuracy and robustness of the final calibration result.

[0082] Based on any of the above embodiments, each parameter combination composed of the reference voltage and the data sampling clock phase is traversed, and the data signal receiving condition under each parameter combination is tested, including:

[0083] The optimal parameter combination obtained by circuit training or the optimal parameter combination determined last time is taken as the starting parameter combination;

[0084] The reference voltage in the starting parameter combination is fixed, and the data sampling clock phase is scanned to the left and right respectively, the data sampling clock phase range in which the data signal can be correctly received under the fixed reference voltage is obtained, and the fixed reference voltage and each data sampling clock phase in the data sampling clock phase range constitute the effective parameter combination under this scanning;

[0085] The fixed reference voltage is changed, and the step of scanning the data sampling clock phase is returned to execute until the scanned data sampling clock phase range under the fixed reference voltage is zero, and all effective parameter combinations are obtained.

[0086] Specifically, the optimal parameter combination obtained by circuit training can be understood as the initial optimal parameter combination determined by transmitting and receiving a standard training sequence in the system power-on or link initial establishment stage, such as the optimal (Vref, deskew). The optimal parameter combination determined last time refers to the optimal parameter combination determined in the last round by executing the signal calibration method described in the embodiment.

[0087] Considering that the link characteristics are usually gradually changed, the optimal parameter combination in the current round is probably located in the neighborhood of the optimal parameter combination obtained by circuit training, or near the last optimal parameter combination, so taking the optimal parameter combination obtained by circuit training or the last optimal parameter combination as the starting parameter combination can greatly reduce the search range of scanning and monitoring, significantly shorten the time to find all effective parameter combinations, and avoid blind search from any corner of the parameter space.

[0088] The reference voltage in the starting parameter combination can be fixed first, and the data sampling clock phase is scanned to the left and to the right respectively to obtain the data sampling clock phase range in which the data signal can be correctly received at the fixed reference voltage, and the fixed reference voltage and each data sampling clock phase in the data sampling clock phase range are constructed into the effective parameter combination in this scan.

[0089] Specifically, the reference voltage of the redundant receiving channel can be set as the reference voltage vref_start in the starting parameter combination. Then, taking the data sampling clock phase deskew_start in the starting parameter combination as the center, the data sampling clock phase is gradually reduced (scanned to the left) respectively, and for each scanned data sampling clock phase, the redundant receiving channel judges whether the current (vref_start, deskew) combination can correctly receive the data signal. If yes, the scanning in the original direction is continued until the deskew makes the reception start to have an error, and the deskew that can correctly receive the signal before the error point is recorded as the lower limit of the data sampling clock phase (deskew_min_at_vref_start) at this voltage. Similarly, the data sampling clock phase is gradually increased (scanned to the right) from deskew_start until the position where the reception starts to have an error is found, so as to determine the upper limit of the data sampling clock phase (deskew_max_at_vref_start).

[0090] Up to now, at vref_start, all the data sampling clock phases located in the interval [deskew_min_at_vref_start, deskew_max_at_vref_start] form parameter pairs with vref_start, and all the parameter pairs are identified as effective parameter combinations.

[0091] After the scanning of vref_start is completed, the reference voltage of the redundant receiving channel is changed to a new value, for example, vref_start+vref_step (scanned upward) or vref_start-vref_step (scanned downward), wherein vref_step is a preset voltage step. Then, the data sampling clock phase scanning process in the first step is returned to be executed to find the effective deskew range at the new Vref value.

[0092] The cycle repeats, for example, starting from vref_start and scanning in both directions, up and down, row by row, to delineate the valid parameter combinations. Until in one direction, when the reference voltage is set to a certain value, the data sampling clock phase range obtained by scanning is zero (i.e., deskew_min is greater than or equal to deskew_max), which indicates that the scan has reached the top or bottom boundary of the valid parameter window, such as Figure 3 the uppermost and lowermost "X" point regions in FIG. 1. At this time, the scan in this direction is terminated. When the scans in both directions are terminated, the entire scanning process is completed.

[0093] Finally, the set of valid parameter combinations identified in the scanning process constitutes the "at least one set of valid parameter combinations" required in step 110, which is fully presented in FIG. 1 as the region composed of all "O" points. Figure 3

[0094] The embodiment searches by fixing the voltage and scanning the phase row by row, which can quickly and completely delineate the entire valid parameter window under the current link state. Compared with the scanning method of traversing all possible parameter combinations, the embodiment greatly shortens the monitoring time and reduces the power consumption of the redundant channel during scanning.

[0095] Based on any of the above embodiments, the optimal parameter combination is determined from the valid parameter combinations according to the position of each valid parameter combination in the valid parameter window, including:

[0096] Calculating the geometric center of the valid parameter window in the reference voltage dimension and the data sampling clock phase dimension;

[0097] Taking the valid parameter combination corresponding to the geometric center as the optimal parameter combination.

[0098] Specifically, the valid parameter window can be understood as the geometric region formed by all valid parameter combinations scanned in step 110 in the two-dimensional parameter coordinate system. As shown in FIG. 1, all points marked with "O" collectively constitute the valid parameter window obtained by this monitoring. The shape and size of the window intuitively reflect the size or health of the signal eye diagram under the current link state. The larger the window, the better the signal quality and the greater the margin. Figure 3

[0099] The geometric center can be understood as the center point of the valid parameter window in the voltage (vertical axis) and timing (horizontal axis) two dimensions. This point is farthest from the boundary of the window (i.e., the critical point from valid to invalid, such as Figure 3 the "X" point in FIG. 1), so it has the largest anti-interference margin. Taking this point as the optimal parameter can enable the receiving end to maintain the lowest bit error rate when facing temperature drift, power noise, signal jitter, etc.​​

[0100] As an optional embodiment, the process of calculating the geometric center is implemented in the following way:

[0101] Among all the valid parameter combinations obtained in step 110, the global minimum value vref_min and the global maximum value vref_max of the reference voltage are found by traversal, and the two values correspond to the lowermost end and the uppermost end of the valid parameter window, respectively. Then, the geometric center vref_best=(vref_min+vref_max) / 2 in the reference voltage dimension is calculated. Figure 3

[0102] On the basis of having calculated vref_best, all valid parameter pairs that satisfy the condition Vref=vref_best are found again by traversing all the valid parameter combinations, and the minimum value deskew_min and the maximum value deskew_max of the data sampling clock phase are determined from them, and the two values correspond to the leftmost end and the rightmost end of the valid parameter window at the horizontal line of vref_best, respectively. Then, the geometric center deskew_best=(deskew_min+deskew_max) / 2 in the data sampling clock phase dimension is calculated. Figure 3

[0103] Through the above two calculation steps, the geometric center point coordinates (vref_best, deskew_best) of the valid parameter window can be obtained. Then, the valid parameter combination corresponding to the geometric center is taken as the optimal parameter combination. That is, the parameter pair (vref_best, deskew_best) calculated in the last step is determined as the optimal parameter combination of the current monitoring and calibration cycle.

[0104] By using the method of calculating the geometric center of the valid parameter window to determine the optimal parameter combination, the embodiment can find the receiver parameters with the maximum voltage margin and timing margin under the current link state in a quantitative and reproducible accurate method, avoiding the uncertainty brought by relying on experience or selecting an arbitrary point in the window.

[0105] Based on any of the above embodiments, the geometric center of the valid parameter window in the reference voltage dimension and the data sampling clock phase dimension is calculated, including:

[0106] determining the maximum reference voltage and the minimum reference voltage among all the valid parameter combinations, and taking the average value of the maximum reference voltage and the minimum reference voltage as the optimal reference voltage, and taking the optimal reference voltage as the geometric center in the reference voltage dimension;

[0107] ​​The maximum data sampling clock phase and the minimum data sampling clock phase in all valid parameter combinations are determined, and the average of the maximum data sampling clock phase and the minimum data sampling clock phase is taken as an optimal data sampling clock phase, and the optimal data sampling clock phase is taken as a geometric center in a data sampling clock phase dimension.

[0108] Specifically, in all valid parameter combinations (Vref, deskew) obtained in step 110, the combinations are traversed to find the maximum reference voltage value vref_max and the minimum reference voltage value vref_min. Then, an optimal reference voltage vref_best=(vref_max+vref_min) / 2 is calculated.

[0109] The vref_best is the midpoint of the valid parameter window in the voltage dimension, and selecting the point as the reference voltage can obtain the maximum voltage margin, so as to better resist the influence of power supply noise and signal attenuation.

[0110] Similarly, in all valid parameter combinations obtained in step 110, the combinations are traversed to find the maximum data sampling clock phase value deskew_max and the minimum data sampling clock phase value deskew_min. Then, an optimal data sampling clock phase deskew_best=(deskew_max+deskew_min) / 2 is calculated.

[0111] The deskew_best is the midpoint of the valid parameter window in the timing dimension, and selecting the point as the sampling clock phase can obtain the maximum timing margin, so as to better resist the influence of signal jitter and clock offset.

[0112] Finally, the combination of the optimal reference voltage and the optimal data sampling clock phase is the optimal parameter combination, that is, the optimal reference voltage vref_best and the optimal data sampling clock phase deskew_best calculated above are combined, that is, the final optimal parameter combination (vref_best, deskew_best) is obtained, which is the geometric center of the valid parameter window shown in FIG. 8, and is theoretically the most robust parameter point in the window. Figure 3

[0113] ​To illustrate this more clearly, let's look at a concrete example. Assume that after scanning and monitoring, the redundant receiving channel determines that all valid parameter combinations cover a reference voltage range of [450mV, 550mV], and a data sampling clock phase range of [-25ps, +25ps]. Accordingly, the optimal reference voltage vref_best = (550mV + 450mV) / 2 = 500mV, and the optimal data sampling clock phase deskew_best = (+25ps + (-25ps)) / 2 = 0ps. Therefore, the determined optimal parameter combination is (500mV, 0ps).

[0114] This embodiment determines the optimal parameter combination by calculating the average value of the effective parameter combination across the entire range of the reference voltage dimension and the data sampling clock phase dimension, providing a computationally simple and highly efficient method for determining optimal parameters. Compared to complex geometric center algorithms that require fitting window shapes, this method only requires simple traversal to find the maximum and minimum values ​​and calculate the average value, greatly reducing the complexity of hardware implementation and the overhead of real-time computation. At the same time, it can quickly and accurately locate parameter points with approximately maximum voltage and timing margins, thereby improving the response speed and practicality of the entire dynamic calibration system while ensuring the finding of robust parameter points.

[0115] Based on any of the above embodiments, applying the optimal parameter combination to the main receiving channel includes:

[0116] After adjusting the reference voltage of the main receiving channel to the optimal reference voltage value in the optimal parameter combination, the data sampling clock phase of the main receiving channel is adjusted to the optimal data sampling clock phase in the optimal parameter combination.

[0117] or,

[0118] After adjusting the data sampling clock phase of the main receiving channel to the optimal data sampling clock phase in the optimal parameter combination, the reference voltage of the main receiving channel is adjusted to the optimal reference voltage value in the optimal parameter combination.

[0119] or,

[0120] Adjust the current parameter combination of the main receiving channel to the optimal parameter combination in one go.

[0121] Here, the optimal reference voltage value is the calculated vref_best, and the optimal data sampling clock phase is the calculated deskew_best. Wherein, Figure 4 This is a schematic diagram of parameter adjustment provided by the present invention, such as... Figure 4 As shown, the current optimal value is the optimal parameter combination, and the previous optimal value is the current parameter combination of the main receiving channel. Figure 4Path 1, pointing from "previous optimal value" to "current optimal value," corresponds to adjusting the reference voltage of the main receiving channel to the optimal reference voltage value in the optimal parameter combination, and then adjusting the data sampling clock phase of the main receiving channel to the optimal data sampling clock phase in the optimal parameter combination. Specifically, the control logic first updates the Vref of the main receiving channel, and then updates its deskew after Vref stabilizes. Since only one parameter is adjusted at a time, the parameter change amplitude is relatively small, and the adjustment process is smoother. This effectively reduces the risk of instantaneous signal decision errors or system instability caused by drastic parameter changes, and improves the stability of the parameter update process.

[0122] As an alternative implementation, the application process can also be reversed, that is, after adjusting the data sampling clock phase of the main receiving channel to the optimal data sampling clock phase in the optimal parameter combination, the reference voltage of the main receiving channel is adjusted to the optimal reference voltage value in the optimal parameter combination.

[0123] This adjustment process corresponds to Figure 4 Path 3 in the diagram. Similar to the methods described above, this method also ensures the smoothness of the update process through sequential adjustments, but in the reverse order. Providing these two different sequences offers flexibility for different hardware implementations. For example, some circuits may be more sensitive to clock phase adjustments; in this case, Vref can be adjusted first, followed by deskew, to stabilize the circuit state under the new voltage margin before performing more critical timing adjustments, thereby reducing the risk of jitter or lockout that may be caused when adjusting the clock phase. Conversely, some circuits may be more sensitive to reference voltage adjustments; in this case, deskew can be adjusted first, followed by Vref, to stabilize the sampling time at the new timing margin center before performing voltage adjustments that may cause fluctuations in the decision maker, thus avoiding bit errors due to insufficient timing margin during voltage adjustments. In other words, the optimal adjustment sequence can be selected based on the specific response characteristics of the hardware to achieve the smoothest transition.

[0124] As another alternative implementation, in order to achieve the fastest adjustment speed, the application process can also be to adjust the current parameter combination of the main receiving channel to the optimal parameter combination all at once.

[0125] This adjustment process corresponds to Figure 4Path 2 in the diagram updates both the Vref and deskew parameters in a single operation, directly switching the parameter combination of the main receiving channel from (vref_current, deskew_current) to (vref_best, deskew_best). This minimizes the parameter adjustment delay, offering a significant advantage in applications with extremely high link calibration response speed requirements. However, it should be noted that if the current parameter combination differs considerably from the optimal combination, a large-scale adjustment in one go may severely impact signal integrity during switching, potentially introducing instantaneous bit errors.

[0126] This embodiment provides designers with the flexibility to balance adjustment speed and switching stability by offering various specific methods for applying optimal parameters to the main receiving channel, such as step-by-step adjustment and one-time adjustment. It allows designers to select the most suitable parameter update strategy according to different application scenarios and performance requirements, thereby optimizing the efficiency of the calibration process while ensuring link reliability, and enhancing the applicability and robustness of the entire signal monitoring and calibration method.

[0127] Based on any of the above embodiments, applying the optimal parameter combination to the main receiving channel includes:

[0128] The data from the primary receiving channel is switched to the redundant receiving channel for transmission.

[0129] After adjusting the current parameter combination of the primary receiving channel to the optimal parameter combination, the data from the redundant receiving channel will be switched back to the primary receiving channel.

[0130] Specifically, there are several ways to apply the optimal parameter combination to the main receiving channel. For example, the parameter combination of the main receiving channel can be updated to the optimal combination all at once, or it can be adjusted step by step. However, although these methods can achieve parameter application, there is a risk of introducing instantaneous signal jitter or errors at the moment of parameter switching, regardless of the switching speed or magnitude. For some systems that require extremely high data transmission zero interruption, such as financial trading systems, real-time industrial control systems, or core server interconnect links, such instantaneous disturbances must be avoided at all costs.

[0131] To address the aforementioned issues, this embodiment provides a preferred, uninterrupted implementation method for applying the optimal parameter combination to the main receiving channel. Specifically, Figure 5 This is a schematic diagram of the switching circuit between the main receiving channel and the redundant receiving channel provided by the present invention. The diagram shows the circuit structure for achieving uninterrupted switching.

[0132] Specifically, first, the data of the main receiving lane is switched to the redundant receiving lane for transmission, which can be understood as a temporary redirection of the data stream. In this embodiment, the switching operation is implemented by a multiplexer (MUX). As shown in FIG. 6, two inputs of the MUX are connected to the output of the main receiving lane (lane0) and the output of the redundant receiving lane (lane0_rd) respectively, and the output of the MUX is the final data output of the physical lane. Figure 5

[0133] Specifically, in the normal working state, the selection control signal of the MUX can be set to "0", so that the data of lane0 is selected and output, that is, the data stream path is: input→lane0→MUX→output. When the parameters of the main receiving lane need to be updated, the control logic changes the selection control signal of the MUX from "0" to "1". At this time, the data stream path is switched to: input→lane0_rd→MUX→output.

[0134] It should be noted that before the switching is performed, it is necessary to ensure that the redundant receiving lane has been configured with a set of valid parameter combinations and can correctly receive the data signal. The valid parameter combination can be the optimal parameter combination obtained in the last calibration, or any valid parameter combination found in the current monitoring process, or even directly set to the optimal parameter combination calculated this time. Since the parameters of the redundant receiving lane are valid, the data stream switching from the main receiving lane to the redundant receiving lane is seamless and will not cause data loss or errors.

[0135] After the data stream is switched to the redundant receiving lane, the main receiving lane is still receiving the same input signal, but its output is not adopted by the downstream logic, which is equivalent to being offline or in the background. At this time, the control logic can write the optimal parameter combination into the corresponding control register of the main receiving lane. Since the main receiving lane does not carry the key real-time data stream at this time, any form of adjustment of its parameters will not cause any impact on the system business.

[0136] After confirming that the parameters of the main receiving lane have been stably set to the optimal parameter combination, the data stream switching back operation is performed. Specifically, the multiplexer (MUX) is controlled again to switch the data stream from the redundant receiving lane back to the calibrated main receiving lane. For example, the selection control signal of the MUX can be reset from "1" to "0". At this time, the main receiving lane continues to undertake the transmission of business data with the updated optimal parameter combination, and the redundant receiving lane is released and can be used for the next round of monitoring and calibration cycle.

[0137] ​The embodiment realizes uninterrupted updating of the main receiving channel parameters by first switching the data stream to the redundant receiving channel seamlessly before updating the main receiving channel parameters, and then switching the data stream back seamlessly after the main receiving channel completes the parameter updating. Since the whole process of parameter adjustment is performed in the offline state of the main receiving channel not carrying the valid data stream, the instantaneous errors or data disturbance possibly introduced due to parameter switching instant are completely avoided, and the absolute continuity and integrity of data transmission are ensured.

[0138] The signal calibration device provided by the present application is described below, and the signal calibration device described below can be correspondingly referred to the signal calibration method described above.

[0139] Based on any of the above embodiments, Figure 6 is a structural schematic diagram of the signal calibration device provided by the present application, as Figure 6 shown, the device comprises:

[0140] The monitoring unit 610 is configured to receive the same data signal by using the redundant receiving channel in parallel with the main receiving channel when the main receiving channel receives the data signal, so that the redundant receiving channel monitors and obtains at least one set of valid parameter combination capable of correctly receiving the data signal.

[0141] The determining unit 620 is configured to determine an optimal parameter combination from the valid parameter combinations according to the positions of the valid parameter combinations in the valid parameter window; the valid parameter window refers to a parameter region capable of correctly receiving the data signal formed by all the valid parameter combinations.

[0142] The calibration unit 630 is configured to apply the optimal parameter combination to the main receiving channel, and calibrate the data signal of the main receiving channel.

[0143] Figure 7 is a structural schematic diagram of the electronic device provided by the present application, as Figure 7 shown, the electronic device can comprise a processor 710, a communications interface 720, a memory 730 and a communications bus 740, wherein the processor 710, the communications interface 720 and the memory 730 complete mutual communication through the communications bus 740. The processor 710 can invoke the logical instructions in the memory 730 to execute the signal calibration method.

[0144] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to related technologies, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0145] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program, the computer program being stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer is able to perform the signal calibration methods provided by the above methods.

[0146] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the signal calibration methods provided by the methods described above.

[0147] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0148] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of software products. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0149] It should be pointed out finally that the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit the same; and although the present application has been described in detail with reference to the foregoing embodiments, it should be appreciated by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features thereof can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A signal calibration method, characterized by, The application comprises the following steps: When receiving a data signal through a main receiving channel, a same data signal is received through a redundant receiving channel parallel to the main receiving channel, so that the redundant receiving channel monitors at least one set of valid parameter combinations capable of correctly receiving the data signal; An optimal parameter combination is determined from each valid parameter combination according to the position of each valid parameter combination in a valid parameter window; the valid parameter window refers to a parameter region capable of correctly receiving the data signal formed by all valid parameter combinations; The optimal parameter combination is applied to the main receiving channel to calibrate the data signal of the main receiving channel; The determination of the valid parameter combination comprises the following steps: Each parameter combination formed by a reference voltage and a data sampling clock phase is traversed, and the data signal receiving condition under each parameter combination is tested; Parameter combinations capable of correctly receiving the data signal are determined as the valid parameter combinations; The determination of the optimal parameter combination from each valid parameter combination according to the position of each valid parameter combination in the valid parameter window comprises the following steps: The geometric center of the valid parameter window in the reference voltage dimension and the data sampling clock phase dimension is calculated; The valid parameter combination corresponding to the geometric center is taken as the optimal parameter combination.

2. The signal calibration method of claim 1, wherein, The traversal of each parameter combination formed by a reference voltage and a data sampling clock phase and the testing of the data signal receiving condition under each parameter combination comprise the following steps: An optimal parameter combination obtained through circuit training or an optimal parameter combination determined last time is taken as a starting parameter combination; The reference voltage in the starting parameter combination is fixed, and the data sampling clock phase is scanned to the left and right respectively, so as to obtain a data sampling clock phase range capable of correctly receiving the data signal under the fixed reference voltage, and each data sampling clock phase in the data sampling clock phase range and the fixed reference voltage form the valid parameter combination under this scanning; The fixed reference voltage is changed, and the scanning of the data sampling clock phase is performed again until the scanned data sampling clock phase range under the fixed reference voltage is zero, and all valid parameter combinations are obtained.

3. The signal calibration method of claim 1, wherein, The calculation of the geometric center of the valid parameter window in the reference voltage dimension and the data sampling clock phase dimension comprises the following steps: The maximum reference voltage and the minimum reference voltage in all valid parameter combinations are determined, and the average value of the maximum reference voltage and the minimum reference voltage is taken as an optimal reference voltage, and the optimal reference voltage is taken as the geometric center in the reference voltage dimension; The maximum data sampling clock phase and the minimum data sampling clock phase in all valid parameter combinations are determined, and the average value of the maximum data sampling clock phase and the minimum data sampling clock phase is taken as an optimal data sampling clock phase, and the optimal data sampling clock phase is taken as the geometric center in the data sampling clock phase dimension.

4. The signal calibration method according to any one of claims 1 to 3, characterized in that, The application of the optimal parameter combination to the main receiving channel comprises the following steps: After the reference voltage of the main receiving channel is adjusted to the optimal reference voltage value in the optimal parameter combination, the data sampling clock phase of the main receiving channel is adjusted to the optimal data sampling clock phase in the optimal parameter combination. or, adjusting the reference voltage of the main receiving channel to the optimal reference voltage value in the optimal parameter combination after adjusting the data sampling clock phase of the main receiving channel to the optimal data sampling clock phase in the optimal parameter combination; or, adjusting the current parameter combination of the main receiving channel to the optimal parameter combination at one time.

5. The signal calibration method according to any one of claims 1 to 3, characterized in that, the application of the optimal parameter combination to the main receiving channel comprises: switching the data of the main receiving channel to the redundant receiving channel for transmission; switching the data of the redundant receiving channel back to the main receiving channel after the current parameter combination of the main receiving channel is adjusted to the optimal parameter combination.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor implements the signal calibration method of any one of claims 1-5 when executing the computer program.

7. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the signal calibration method of any one of claims 1-5.

8. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the signal calibration method of any one of claims 1-5.

Images