IQ skew estimation method and apparatus

Abstract

The application discloses an IQ skew estimation method, which uses a dynamic scanning unit to estimate a skew_index value corresponding to an IQ skew value according to statistical correlation value information of I and Q data, uses a dynamic tracking unit to track and adjust the skew_index of the IQ skew value, and finally outputs a skew estimation value according to the skew_index. The application also discloses an IQ skew estimation method, which only uses a dynamic scanning unit to estimate a skew_index value corresponding to an IQ skew value, and outputs a skew estimation value according to the skew_index. The application also discloses an IQ skew estimation method, which only uses a dynamic tracking unit to estimate a skew_index value corresponding to an IQ skew value, and outputs a skew estimation value according to the skew_index. The application also discloses corresponding IQ skew estimation devices.

Description

Technical Field

[0001] This invention relates to the field of optical fiber communication technology, and more specifically to an IQ skew estimation method and apparatus in a digital coherent receiver. Background Technology

[0002] In digital coherent optical communication systems, with the gradual application of higher-order modulation formats and higher symbol rates, coherent optical systems are becoming increasingly sensitive to transceiver defects. One significant factor affecting the performance of high-speed transmission systems is the time deviation between the in-phase / quadrature (IQ) branches of the transceiver, known as IQ skew. The impact of skew on performance degradation varies depending on the modulation format; in higher-order modulation formats, the impact of skew is more pronounced. Therefore, accurate estimation and compensation of skew are crucial for improving the performance of high-speed coherent optical communication systems.

[0003] To reduce transmission performance loss caused by IQ skew, existing technologies include methods such as using special transmission sequences and instrument-assisted measurements, methods combining fiber dispersion estimation, and equalization-type estimation algorithms to estimate and compensate for IQ skew. For example... Figure 1 The digital signal processing (DSP) flowchart at the receiver end of a digital coherent fiber optic communication system shows that the IQ calibration module, dispersion compensation module, and polarization demultiplexing module are all used to estimate and compensate for the receiver's IQ skew. The joint estimation algorithm uses the IQ calibration module and dispersion compensation module to jointly estimate and compensate for IQ skew and dispersion. This algorithm uses a two-dimensional step-scan approach, resulting in high computational complexity and dependence on subsequent modules. Equalization-based estimation algorithms use the polarization demultiplexing module to compensate for IQ skew, calculating the IQ skew value through equalization coefficients. These algorithms are highly complex and have significant hardware implementation costs. Methods involving transmitting special sequences and instrument-assisted measurements have narrow applicability. Furthermore, IQ skew estimation should be performed as close as possible to the receiver's input side, and accurate compensation of the IQ skew should be achieved as quickly as possible to minimize the impact on subsequent modules. Figure 1 An IQ calibration module is a suitable choice, and this application performs IQ skew estimation compensation within the IQ calibration module. Therefore, to overcome the performance degradation problem caused by IQ skew in signal transmission, a highly accurate and easily implemented IQ skew estimation method and apparatus are needed. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an IQ skew estimation method and apparatus, which can reduce computational complexity when estimating IQ skew, while achieving large-scale, high-precision estimation.

[0005] To achieve the above objectives, according to one aspect of the present invention, an IQ skew estimation method is provided. Based on the statistical correlation information of I-channel and Q-channel data pairs, a dynamic scanning unit is used to estimate the skew-index value corresponding to the IQ skew value. Then, a dynamic tracking unit is used to track and adjust the skew-index of the IQ skew value. Finally, the skew estimation value is output based on the skew-index.

[0006] In one embodiment of the present invention, estimating the skew-index value corresponding to the IQ skew value using a dynamic scanning unit specifically includes:

[0007] The initial scan direction is determined based on the sign of the initial relevant information mean. This continues until the signs of the (M-1)th and Mth scan means are the same and different from the initial sign. Then, it is determined whether the signs of the 1st and 2nd scan means are the same to decide whether a reverse scan is needed at the initial skew value. The scan ends when the signs of the 1st and 2nd scan means are the same, and the signs of the (M-1)th and Mth scan means are the same and opposite to the signs of the 1st and 2nd scan means. At the end of the scan, the skew_index value corresponding to the relevant information mean with the smallest absolute value is obtained.

[0008] In one embodiment of the present invention, when using a dynamic tracking unit to track and adjust the skew-index of the IQ skew value:

[0009] The direction prediction unit predicts the starting direction of each dynamic tracking step based on the relevant mean information compensated with skew_index0 within the waiting window. If the starting direction is positive, the absolute values ​​of the relevant mean compensated with skew_index0+1 and skew_index0 are compared. If the current absolute value is smaller, the current dynamic tracking ends and skew_index0+1 is output as the estimated value. If the current absolute value is larger, the absolute values ​​of the relevant mean compensated with skew_index0-1 and skew_index0 are compared, and the IQskew estimate corresponding to the smaller absolute value is output. The entire tracking process is the same when the starting direction is negative, except that the direction is reversed, where skew_index0 is the initial value of the dynamic tracking IQ skew.

[0010] According to another aspect of the present invention, an IQ skew estimation method is also provided, which uses a dynamic scanning unit to estimate the skew-index value corresponding to the IQ skew value based on the statistical correlation information of the I-channel and Q-channel data pairs, and outputs the skew estimation value based on the skew-index.

[0011] According to another aspect of the present invention, an IQ skew estimation method is also provided, which uses a dynamic tracking unit to estimate the IQ skew value based on the statistical correlation information of I-channel and Q-channel data to obtain the skew-index, and outputs the skew estimation value based on the skew-index.

[0012] In one embodiment of the present invention, the method further includes anomaly monitoring, specifically: obtaining the absolute value information of the difference between the relevant values ​​of adjacent statistical windows; if the absolute value of the current difference exceeds k times the absolute value of the previous difference, it indicates that there is an abnormal jump in IQskew at this time, and the software selects whether to reset and restart the scanning and tracking process.

[0013] In one embodiment of the present invention, the step of statistically analyzing correlation information based on I-channel and Q-channel data specifically includes:

[0014]

[0015] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, and K is the weight information.

[0016] In one embodiment of the present invention, when the interference in the channel is lower than a preset value, the method for directly calculating the IQ skew value is as follows:

[0017]

[0018] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, K is the weight information, Δω is the frequency offset information, and Δτ is the IQ skew value.

[0019] According to another aspect of the present invention, an IQ skew estimation device is also provided, including an IQ skew compensation unit, other damage compensation unit, relevant information statistics unit, and IQ skew estimation update unit, wherein the IQ skew estimation update unit estimates the IQ skew using the method described above.

[0020] In one embodiment of the present invention, the relevant information statistics unit uses the following expression to count relevant value information:

[0021]

[0022] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, and K is the weight information.

[0023] In summary, the technical solutions conceived by this invention have the following beneficial effects compared with the prior art:

[0024] (1) The method of the present invention has a wide range of applications and can be used for skew estimation of various transmission rate signals in various modulation formats such as QPSK, 8QAM, 16QAM, and 64QAM;

[0025] (2) The estimation and tracking accuracy of the method of the present invention can reach 1 / 32 of the sample symbol period, which is highly accurate. The skew_index estimation and tracking range can reach -32 to 31, and the adjustment range is the two adjacent sample symbol periods, which has a large skew estimation range.

[0026] (3) Compared with equalization filtering techniques, the method of the present invention has low computational complexity. Compared with the simultaneous calculation of multiple branches, this technique only uses one path for estimation and tracking.

[0027] (4) The method of the present invention uses symbol to determine the scanning direction and scanning stop position in the dynamic scanning process, and focuses on the detection of key areas. The region decision improves the scanning accuracy. The dynamic tracking process uses probability prediction, adjusts the direction to reduce the number of adjustments, reduces the tracking time and dynamically reduces power consumption, and can track the changing IQ skew.

[0028] (5) The method of the present invention has the ability to monitor skew anomalies, and can automatically restart the estimation process for large anomaly jumps without using slow tracking and adjustment, thus reducing recovery time; scanning, tracking and anomaly monitoring can be used independently to enhance flexibility, and can be used in combination to improve stability. Attached Figure Description

[0029] Figure 1 This is a flowchart of the digital signal processing (DSP) at the receiver end of a digital coherent optical fiber communication system.

[0030] Figure 2 This is a schematic diagram illustrating the relationship between correlation values ​​and skew for different modulation formats; where... Figure 2 (a) shows the QPSK modulation format. Figure 2 (b) is the 8QAM modulation format. Figure 2 (c) represents the 16QAM modulation format;

[0031] Figure 3 This is a block diagram of IQ skew estimation and compensation in an embodiment of the present invention;

[0032] Figure 4 This is a flowchart of the IQ skew estimation and update process in an embodiment of the present invention;

[0033] Figure 5 This is a flowchart of the IQ skew dynamic scanning process in an embodiment of the present invention;

[0034] Figure 6 This is a schematic diagram of IQ skew dynamic scanning in an embodiment of the present invention;

[0035] Figure 7 This is a graph showing the changes in the IQ skew dynamic scan estimation results in an embodiment of the present invention;

[0036] Figure 8 This is a flowchart of the IQ skew dynamic tracking process in an embodiment of the present invention;

[0037] Figure 9 This is a schematic diagram of IQ skew dynamic tracking in an embodiment of the present invention;

[0038] Figure 10 This is a graph showing the changes in the IQ skew dynamic tracking estimation results in an embodiment of the present invention;

[0039] Figure 11 This is a graph showing the changes in the joint estimation results of IQ skew dynamic scanning and dynamic tracking in an embodiment of the present invention;

[0040] Figure 12 This is a flowchart of IQ skew abnormal alarm and recovery in an embodiment of the present invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0042] This application provides a method for calculating IQ-related information for IQ skew estimation and tracking. Based on the statistical correlation information of I-channel and Q-channel data, the skew-index value corresponding to the IQ skew value is estimated using a dynamic scanning unit. Then, the skew-index is adjusted by tracking the IQ skew value using a dynamic tracking unit. Finally, the skew estimate is output based on the skew-index.

[0043] Specifically, the step of estimating the skew-index value corresponding to the IQ skew value using a dynamic scanning unit includes:

[0044] The initial scan direction is determined based on the sign of the initial relevant information mean. This continues until the signs of the (M-1)th and Mth scan means are the same and different from the initial sign. Then, it is determined whether the signs of the 1st and 2nd scan means are the same to decide whether a reverse scan is needed at the initial skew value. The scan ends when the signs of the 1st and 2nd scan means are the same, and the signs of the (M-1)th and Mth scan means are the same and opposite to the signs of the 1st and 2nd scan means. At the end of the scan, the skew_index value corresponding to the relevant information mean with the smallest absolute value is obtained.

[0045] When using the dynamic tracking unit to track and adjust the skew-index of the IQ skew value, the direction prediction unit predicts the starting direction of each dynamic tracking based on the relevant mean information compensated with skew_index0 within the waiting window. If the starting direction is positive, the absolute values ​​of the relevant mean compensated with skew_index0+1 and skew_index0 are compared. If the current absolute value is smaller, the dynamic tracking ends and skew_index0+1 is output as the estimated value. If the current absolute value is larger, the absolute values ​​of the relevant mean compensated with skew_index0-1 and skew_index0 are compared, and the IQ skew estimate corresponding to the smaller absolute value is output. The entire tracking process is the same when the starting direction is negative, except that the direction is reversed, where skew_index0 is the initial value of the IQ skew for dynamic tracking.

[0046] Furthermore, based on the statistical correlation information of the I-channel and Q-channel data, the skew-index value corresponding to the IQ skew value can be estimated using the dynamic scanning unit, and the skew estimate value can be output based on the skew-index; or, based on the statistical correlation information of the I-channel and Q-channel data, the skew value can be estimated using the dynamic tracking unit to obtain the skew-index, and the skew estimate value can be output based on the skew-index.

[0047] In addition, anomaly monitoring can be performed during the estimation and tracking process. The specific method is as follows: obtain the absolute value information of the difference between the relevant values ​​of adjacent statistical windows. If the absolute value of the current difference exceeds k times the absolute value of the previous difference, it indicates that there is an abnormal jump in IQskew at this time. The software can choose whether to reset and restart the scanning and tracking process.

[0048] Furthermore, based on the I-path and Q-path data, statistical correlation information is obtained, specifically including:

[0049]

[0050] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, and K is the weight information.

[0051] Furthermore, when the interference in the channel is below a preset value, the method for directly calculating the IQ skew value is as follows:

[0052]

[0053] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, K is the weight information, Δω is the frequency offset information, and Δτ is the IQ skew value;

[0054] The K value in the relevant information statistical expression is adjusted according to the magnitude of the monitored phase imbalance. Specifically: if the IQ phase imbalance damage in the system is greater than the preset value, then the correlation between the I and Q paths is strong, and K is adjusted to 1; when the IQ phase imbalance damage in the system is less than the preset value or has been compensated, K is selected as 3; the default selection is K=2.

[0055] When the interference in the channel is lower than the preset value, the specific methods for directly calculating the IQ skew value are as follows: the estimated value can be obtained by direct inverse mapping based on the approximate linear relationship of the calculation expression, or by using a lookup table and inverse trigonometric function to calculate the estimated value, or by using software to calculate the estimated value.

[0056] "Below the preset value" generally refers to a situation where there is less interference, i.e., the channel is relatively ideal. Specifically, a preset value can be set according to actual needs. When the interference is below the preset value, the channel is considered unideal, and vice versa.

[0057] The technical solution of this invention will be further described in detail below. This invention utilizes the statistical correlation information of the I-channel and Q-channel data pairs to further process and analyze the correlation values ​​to obtain the IQ skew estimate. After statistical averaging, the statistical expression for the correlation information is as follows:

[0058]

[0059] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, K is the weight information, Δω is the frequency offset information, and Δτ is the IQ skew value.

[0060] In the absence of other interference, the relevant statistical expression approximates a sine function, and the estimated range of Δτ is half a monotonic sine cycle. Within a preset range, the greater the deviation of the correlation value from the zero axis, the greater the residual skew value in the signal.

[0061] However, in practice, it is particularly important to consider the impact of phase imbalance between the I and Q paths on the accuracy of IQ skew estimation. The K value in the relevant information statistical expression should be adjusted based on the magnitude of the monitored phase imbalance. The K value can be configured according to the magnitude of the residual IQ phase imbalance in the system to meet the IQ skew estimation requirements of different scenarios. If the IQ phase imbalance in the system is significant, the correlation between the I and Q paths is strong, and the direct correlation value ∑I between the two signals is high. n ·Q nA large value will affect the IQskew estimation result. In this case, when using relevant information statistics expressions to estimate IQskew, K can be adjusted to 1. When K=1, the product term of I and Q is reduced, thus reducing the impact of phase correlation between the I and Q paths. When the IQ phase imbalance in the system is small or has been compensated, a value of 3 can be selected to increase the relevant information statistics result, making its resolution more obvious when passing through the zero axis, which is more conducive to the estimation of IQskew. Considering the above, the default value is K=2. This scheme is generally applicable to situations where the frequency deviation of the signal is greater than 300MHz. When the frequency deviation is large, the correlation value and the correspondence between IQskew are more significant.

[0062] This application provides a method for IQ skew estimation, such as... Figure 3 As shown, it includes an IQ skew compensation unit 31, an other damage compensation unit 32, an relevant information statistics unit 33, and an IQ skew estimation update unit 34. The IQ skew estimation update unit 34 updates the relevant information again by the relevant information statistics unit 33 after the IQ skew estimated value remaining in the relevant information estimation update signal obtained by the relevant information statistics unit 33 is compensated by the IQ skew compensation unit 31, and the relevant information is updated by the relevant information statistics unit 33. Iterative estimation and compensation are performed to gradually make the estimated value approach the true value.

[0063] The relevant information statistics unit 33 uses the following expression to count relevant value information:

[0064]

[0065] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, and K is the weight information;

[0066] The IQ skew estimation update unit 34 updates the IQ skew estimate using the IQ skew estimation method.

[0067] Specifically, when using relevant information statistical expressions for skew estimation, this application primarily relies on the changing trends of the skew, using the patterns of change in its sign and magnitude to approximate and estimate the IQ skew. For example... Figure 4As shown in the flowchart of the IQ skew estimation update, the IQ skew estimation update unit 34 provided in this application consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, a dynamic tracking unit 43, an anomaly monitoring unit 44, and a skew estimation output 45. The estimation schemes can be divided into four types: (1) using a dynamic scanning unit to estimate IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, and a skew estimation output 45; (2) using a dynamic tracking unit to estimate IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic tracking unit 43, and a skew estimation output 45; (3) using a dynamic scanning unit and a dynamic tracking unit to estimate IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, a dynamic tracking unit 43, and a skew estimation output 45; (4) using a dynamic scanning unit and an anomaly monitoring unit to estimate IQ skew, using a dynamic tracking unit and an anomaly monitoring unit to estimate IQ skew, and using a dynamic scanning unit, a dynamic tracking unit, and an anomaly monitoring unit to estimate IQ skew, which can be used as a supplement to schemes (1), (2), and (3).

[0068] Example 1

[0069] like Figure 1 As shown, this technology is applied to the IQ calibration module at the receiving front end in the DSP flowchart. For input signals in X-polarization or Y-polarization states, the IQ calibration module processes the IQ impairments present in the two polarization states, including IQ amplitude imbalance impairment, IQ phase imbalance impairment, and IQ skew impairment. This technical solution addresses IQ skew impairment by utilizing the statistical correlation information of the I-channel and Q-channel data pairs to further process and analyze the relevant information to obtain an estimated IQ skew value.

[0070] Throughout the communication process, if the receiver only considers the IQ skew impairment factor, the in-phase component I of the coherent received signal... n and orthogonal components Q n The expression is as follows:

[0071]

[0072]

[0073] The expressions for statistical correlation information of I-channel and Q-channel data are shown below:

[0074]

[0075] Δτ can be considered a constant value over a period of time. After statistical averaging, the relevant statistical expression for the sine function is as follows:

[0076]

[0077] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, K is the weight information, Δω is the frequency offset information, and Δτ is the IQ skew value.

[0078] In the absence of other interference, the relevant statistical expression approximates a sine function, and the estimated range of Δτ is close to half a sine cycle. Within a certain range, the greater the deviation of the correlation value from the zero axis, the greater the residual skew value in the signal.

[0079] However, in practice, it is particularly important to consider the impact of phase imbalance between the I and Q paths on the accuracy of IQ skew estimation. The K value in the relevant information statistical expression should be adjusted according to the magnitude of the monitored phase imbalance. If the IQ phase imbalance in the system is significant, the correlation between the I and Q paths is strong, and the direct correlation value ∑I between the two signals will be high. n ·Q n A large value will affect the IQ skew estimation result. In this case, when using relevant information statistics expressions for IQ skew estimation, K can be adjusted to 1. When K=1, the product term of I and Q is reduced, thus reducing the phase correlation between the I and Q paths. When the IQ phase imbalance in the system is small or has been compensated, a value of 3 can be selected to increase the relevant information statistics result, making it more resolute when passing through the zero axis, which is more conducive to IQ skew estimation. Considering all of the above, the default value is K=2. The K value can be configured by the DSP according to the magnitude of the residual IQ phase imbalance value to meet the IQ skew estimation requirements of different scenarios. This scheme is usually used when the frequency deviation in the signal is greater than 300MHz. When the frequency deviation is large, the correlation value and the IQ skew correspond more significantly.

[0080] Figure 2 The relationship curves between the relevant information statistics and IQskew of the IQ and IQ signals are presented for three modulation formats: QPSK, 8QAM, and 16QAM. The impact of different K values ​​is also explained. Figure 2 (a) is the QPSK modulation format. Figure 2 (b) is an 8QAM modulation format. Figure 2 (c) is a 16QAM modulation format. In this format, different magnitudes of IQ skew Δτ are introduced between the I and Q signals when the signal passes through the channel. From... Figure 2 It can be seen that under low-order modulation, especially QPSK modulation, the relationship between the two is approximately a straight line within a certain range. The higher the modulation order, the more obvious the curve relationship presented by the relevant information statistical expression provided in this application becomes.

[0081] In specific stable transmission environments, QPSK can directly obtain an approximate estimate of Δτ by inverse mapping using relevant information statistics based on the linear relationship between the two. For other modulations, an approximate estimate of Δτ can be obtained by looking up tables or using inverse trigonometric functions or by software calculation. However, in typical usage scenarios, signals are greatly affected by factors such as nonlinear optical fiber channels and imperfections in optical devices. The types and degrees of impairments introduced into the signal are numerous and vary. It is difficult for relevant information statistics and the skew relationship to perfectly match the aforementioned expression. Directly using approximate expressions for inverse mapping or looking up tables yields poor results.

[0082] When using relevant information statistical expressions for skew estimation, this application primarily relies on the changing trends of the skew, using the patterns of change in its sign and magnitude to approximate the IQ skew. For example... Figure 3 As shown, the system includes a relevant information statistics unit 33, an IQ skew estimation and update unit 34, an IQ skew compensation unit 31, and other impairment compensation units 32. Specifically: the relevant information statistics unit 33 is used to acquire relevant statistical information of the I-path and Q-path; the IQ skew estimation and update unit 34 is used to estimate the current IQ skew value in the system based on the statistical information provided by the relevant information statistics unit 33; the IQ skew compensation unit 31 is used to compensate for IQ skew impairments in the system based on the IQ skew estimate fed back by the IQ skew estimation and update unit 34; and the other impairment compensation unit 32 is used to compensate for other performance impairments of the I-path and Q-path besides IQ skew, namely DC bias, amplitude imbalance impairment between the I-path and Q-path, and phase imbalance impairment between the I-path and Q-path. The residual IQ skew estimate in the updated signal estimated by the IQ skew estimation and update unit 34 is compensated by the IQ skew compensation unit 31, and then iteratively estimated and compensated again, gradually bringing the estimated value closer to the true value. In this process, the method and apparatus provided in this application are implemented through 33 relevant information statistics unit and 34 IQ skew estimation and update unit.

[0083] like Figure 4As shown in the flowchart of the IQ skew estimation update, the IQ skew estimation update unit 34 provided in this application consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, a dynamic tracking unit 43, an anomaly monitoring unit 44, and a skew estimation output 45. The estimation schemes can be divided into four types: (1) using a dynamic scanning unit to estimate IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, and a skew estimation output 45; (2) using a dynamic tracking unit to estimate IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic tracking unit 43, and a skew estimation output 45; (3) using a dynamic scanning unit and a dynamic tracking unit to estimate IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, a dynamic tracking unit 43, and a skew estimation output 45; (4) using a dynamic scanning unit and an anomaly monitoring unit to estimate IQ skew, using a dynamic tracking unit and an anomaly monitoring unit to estimate IQ skew, and using a dynamic scanning unit, a dynamic tracking unit, and an anomaly monitoring unit to estimate IQ skew, which can be used as a supplement to schemes (1), (2), and (3).

[0084] Example 2

[0085] This embodiment is a specific implementation of scheme (1) in embodiment 1, such as Figure 4 The scheme (1) shown uses a dynamic scanning unit to estimate the IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, and a skew estimation output 45. The dynamic scanning scheme can be preferentially applied to scenarios where the skew value is large and IQ skew needs to be quickly acquired for compensation.

[0086] In this embodiment, the IQ-related information acquisition unit 41 is used to obtain the IQ-related information for each statistical window. Figure 5The correlation mean calculation unit 51 averages the relevant information from unit 41, using this average as a condition for adjusting the IQskew estimation direction of each dynamic scan unit 42. The scan direction decision unit 52 determines the starting direction of the scan estimation based on the sign of the first correlation mean. Taking a forward scan as an example, at the end of each statistical window, the skew estimation output unit 60 feeds back the estimated value skew_index+1 to the IQskew compensation unit 31, and each feedback estimate is incremented by one based on the previous one. At the same time, unit 54 determines whether the condition that the signs of the mean values ​​of the (M-1)th and Mth scans are the same and negative is met. Until the condition in unit 54 is met, unit 56 checks if the signs of the first and second means are the same and positive. If the condition is met, the skew_index corresponding to the mean with the smallest absolute value in this entire dynamic scan process is used as the final dynamic scan result and fed back to the IQ skew compensation unit 31 by the skew estimation output unit 45. If the condition in unit 56 is not met, the skew estimation output unit 60 feeds back the initial skew_index-1 to the IQ skew compensation unit 31. Similarly, each feedback estimate is reduced by one from the previous one to achieve reverse scan estimation. At the same time, the first and second means are updated until the condition in unit 56 is met. Then, the skew estimation output unit 45 outputs the skew_index corresponding to the mean with the smallest absolute value in the final scan process to the IQ skew compensation unit 31. The reverse scan process is the same as the forward scan, except that the direction is reversed.

[0087] The implementation process of the dynamic scanning unit 42 is as follows: Figure 5As shown. The IQ-related information receiving unit 50 receives relevant information statistics from the relevant information statistics unit 33. In hardware implementation, these statistics are typically achieved using an IIR low-pass filter, whose average output value gradually approaches the mean as the data volume increases. Therefore, a certain amount of time or data is required for the average output of the relevant information statistics unit 33 to stabilize. During the implementation of the dynamic scanning unit 42, the software ensures that the average output of 33 reaches a relatively stable state within the statistical window specified by the data volume, based on the configurable number of data beats in the statistical window. Typically, at the beginning of the statistical window, the relevant information statistics are gradually increasing or decreasing; at the end of the statistical window, the relevant information statistics tend to stabilize. At the end of each statistical window, the dynamic scanning unit 42 updates the IQ skew estimate skew_index and transmits it to the IQ skew compensation unit 31. In this embodiment, the signal sample time interval is T. When compensating for the skew, the IQ skew compensation unit 31 upsamples the signal samples by a factor of 32. It selects filter coefficients based on the input skew_index value, completes the compensation using an FIR filter, and then downsamples the output. Here, skew_index represents the magnitude and sign of the skew, and its value range is related to the signal upsampling factor. The skew_index varies from -32 to 31, and its adjustment range is the interval between two adjacent samples. The adjustment precision during compensation is T / 32.

[0088] The correlation value mean calculation unit 51 selects only a portion of relevant statistical values ​​from the end of a statistical window for storage. For example, within the last 1 / 4 width, 16 points are retained at equal intervals along the data flow. The selection position and number are not limited in this application, but the number of points should not be too large, as long as the design accuracy and hardware overhead are met. Before the end of the statistical window, the correlation value mean calculation unit 51 recalculates the mean of the stored data and outputs it, outputting once for each statistical window.

[0089] If M window scans are performed, the scanning process saves the average values ​​output by the 1st, 2nd, (M-1)th, and Mth correlation value calculation unit 51 for subsequent scan process decisions. The 1st average value determines the scan direction, i.e., whether the IQ skew estimate skew_index moves towards the positive increment direction or the negative decrement direction at the initial stage of the scan, as shown in the scan direction decision unit 52. The scan ends when the signs of the 1st and 2nd average values ​​are the same, and the signs of the (M-1)th and Mth average values ​​are the same but opposite to the signs of the 1st and 2nd average values, as shown in decision units 54 and 56. There are five scan processes based on the sign of the 1st average value, as detailed in the following scan processes: Figure 5The IQ skew dynamic scanning flowchart is shown below. The five scanning processes are briefly described as follows: When the first mean is positive, the initial scanning direction of skew_index is positive plus 1. After compensation by units 54, 55, 60, and compensation unit 31, another scanning process is performed until unit 54 produces two consecutive negative mean values ​​for the (M-1)th and Mth times. If unit 56 determines that both the first and second mean values ​​are positive, then the skew_index corresponding to the mean with the smallest absolute value is taken as the dynamic scanning result, such as... Figure 6 The IQ skew dynamic scanning diagram (top) shows that when the first mean is positive, the initial scanning direction of skew_index is positive and incremented by 1. After compensation by units 54, 55, 60 to compensation unit 31, another scanning process is performed until unit 54 shows two consecutive negative mean values ​​for the (M-1)th and Mth times. If the first mean is positive and the second mean is zero or negative, the skew_index corresponding to the first mean is decremented by one. After compensation by units 56, 57, 58, 60 to compensation unit 31, another scanning process is performed, and the first and second mean values ​​are updated until two consecutive positive mean values ​​appear. The skew_index corresponding to the mean with the smallest absolute value is taken as the dynamic scanning result. Figure 6 The IQ skew dynamic scanning diagram (middle) shows that when the first mean is 0, the initial scanning direction of skew_index is positive and incremented by 1. After compensation at units 54, 55, 60, and 31, another scanning process is performed until two consecutive negative mean values ​​appear (M-1th and Mth times). At this point, the skew_index corresponding to the first mean is decremented by 1. After compensation at units 56, 57, 58, 60, and 31, another scanning process is performed, updating the first and second mean values ​​until two consecutive positive mean values ​​appear. The skew_index corresponding to the mean with the smallest absolute value is taken as the dynamic scanning result, as shown below. Figure 6The diagram below illustrates the IQ skew dynamic scanning process. When the first mean is negative, the skew_index is decremented by 1. After compensation by the compensation unit 31, the scanning process is repeated until two consecutive positive mean values ​​appear (M-1 and M). If both the first and second mean values ​​are negative, the skew_index corresponding to the mean with the smallest absolute value is taken as the dynamic scanning result. When the first mean is negative, the skew_index is decremented by 1. After compensation by the compensation unit 31, the scanning process is repeated until two consecutive positive mean values ​​appear (M-1 and M). If the first mean is negative and the second mean is zero or positive, the skew_index corresponding to the first mean is incremented by 1, a reverse scan is performed, and the first and second mean values ​​are updated until two consecutive negative mean values ​​appear. The skew_index corresponding to the mean with the smallest absolute value is taken as the dynamic scanning result.

[0090] The simulation conditions were QPSK modulation format, baud rate of 30.07 GHz, 1.33x upsampling, sampling factor of 32, with -2 ps IQ skew impairment added to the signal, a scanning window of 2^20 data beats, and a frequency offset of 1 GHz. The simulation results are as follows: Figure 7 As shown in the dual-axis graph, the left ordinate represents the mean of IQ-related information, and the right ordinate represents the estimated IQ skew value, skew_index. The first mean is negative, and the scan proceeds in the reverse direction, decreasing by 1. When two consecutive positive mean values ​​appear, the reverse scan ends, and it is determined that both the first and second mean values ​​are negative. At this point, the scanning process ends, and the skew_index = -3 corresponding to the mean with the smallest absolute value is taken as the dynamic scan result. Subsequent compensation unit 31 maintains skew_index = -3 to compensate the signal. The skew value corresponding to skew_index = -3 is -2.3 ps, the minimum adjustment accuracy during estimation is 0.78 ps, and the simulation error for this dynamic scan is 0.34 ps.

[0091] IQskew=skew_index / (1.33×30.07e 9 ×32)×10 12 ≈-2.34ps.

[0092] Example 3

[0093] This embodiment is a specific implementation of scheme (2) in embodiment 1, such as Figure 4 The scheme (2) shown uses a dynamic tracking unit to estimate IQ skew, which consists of an IQ-related information acquisition unit 41, a dynamic tracking unit 43, and a skew estimation output 45. The dynamic tracking scheme is preferably applied to scenarios where IQ skew is small or changes and requires long-term monitoring.

[0094] In this embodiment, the IQ-related information acquisition unit 41 is used to obtain the IQ-related information for each statistical window. Figure 8 The correlation mean calculation unit 81 averages the relevant information from unit 41, using this average as a condition for adjusting the IQskew estimation direction of each dynamic tracking unit 43. The IQskew estimation value skew_index0 fed back to the IQskew compensation unit 31 within the waiting window remains unchanged. The direction prediction unit 82 statistically analyzes the correlation mean information within the waiting window to predict the starting direction of each dynamic tracking iteration. If the starting tracking direction is positive, D in the adjustment unit 83 equals +1; otherwise, it is -1. Taking a positive initial tracking direction as an example, if D in unit 83 is +1, then skew_index0+1 is fed back to the IQskew compensation unit 31. Subsequently, comparison unit 84 compares the absolute value of the correlation mean after compensation with skew_index0+1 with the absolute value of the correlation mean after compensation with skew_index0. If the current absolute value is smaller, the current dynamic tracking ends, and the skew_index0+1 corresponding to the current absolute value is fed back to unit 31 as the final estimate for this tracking. If the current absolute value is larger, skew_index0-1 is fed back to unit 31 as the estimate, and the absolute value at this time is compared with the absolute value of the correlation mean after compensation with skew_index0. The IQskew estimate corresponding to the smaller absolute value is fed back as the final estimate for this dynamic tracking. Before the next dynamic tracking begins, IQskew compensation is performed using the final estimate for this tracking. The entire tracking process is the same for a negative initial direction, except that the direction is reversed.

[0095] The implementation process of dynamic tracking unit 43 is as follows: Figure 8As shown. The IQ-related information receiving unit 80 receives relevant information statistics values ​​from the relevant information statistics unit 33. During the implementation of the dynamic tracking unit 43, the software will ensure that the average value output by 33 can reach a relatively stable state within the statistical window specified by the data beat count of the configurable distribution statistical window according to the usage scenario. During the implementation of the dynamic tracking unit 43, the number of waiting statistical windows can be configured selectively. During this period, the IQ skew estimate transmitted to the IQ skew compensation unit 31 remains unchanged. At the end of each waiting window, the direction prediction unit 82 will record the mean sign and magnitude output by the correlation value mean calculation unit 81. The correlation value mean calculation unit 81 has the same function as the correlation value mean calculation unit 51, and 80, 81 and 50, 51 can be reused. When all waiting statistical windows end, the tracking direction is determined by the direction prediction unit 82. The direction prediction unit 82 determines the direction of the next IQ skew value adjustment—whether it's a positive increment or a negative decrement—based on a comprehensive analysis of the recorded mean sign and magnitude, as shown in the skew adjustment unit 83. Here, skew_index = skew_index + D, where D is the adjustment direction, 1 represents an increment of skew, and -1 represents a decrement of skew. The analysis and prediction method is not limited in this application; any method can be used as long as it meets the requirements of prediction accuracy and hardware overhead.

[0096] Comparison unit 84 uses the mean of relevant information obtained from the last waiting window, Mean0, and the estimated IQ skew value, skew_index0, as references for tracking and adjustment.

[0097] If the initial tracking direction is skew_index + 1, and D = 1, then skew_index1 = skew_index0 + 1 is passed to the IQ skew compensation unit 31. After passing through a statistical window, the absolute values ​​of the current relevant information mean Mean1 and Mean0 are compared in unit 84. If |Mean1| ≤ |Mean0|, then skew_index1 = skew_index0 + 1 is fed back to the compensation module 31, as shown in units 84, 86, and 89. This indicates that the dynamic tracking ends, and the tracking adjustment count is 1. Figure 9 The IQskew dynamic tracing diagram (left) shows points A* and B*; if |Mean1|>|Mean0|, as shown... Figure 9 The IQ skew dynamic tracking diagram (left) shows points A and B. Then, using skew_index0 as the baseline, the tracking proceeds in reverse for another window, with D = -1. Therefore, skew_index... -1 =skew_index0-1 is fed back to compensation module 31, as shown in units 85, 88, and 89. After a statistical window, the mean value of the current relevant information is compared in unit 84. -1The absolute value of Mean0 is used to track the adjustment count of 2, such as... Figure 9 IQ skew dynamic tracking diagram (left) showing points A and C. If |Mean -1 If |≤|Mean0|, then skew_index -1 =skew_index0-1 is fed back to compensation module 31, as shown in units 84, 86, and 89, indicating the end of this dynamic tracking; if |Mean -1 If |>|Mean0|, then skew_index0 will be fed back to compensation module 31, as shown in units 85, 87, and 89, which means the dynamic tracking ends.

[0098] If the initial tracking direction is skew_index minus one, and D = -1, then the skew_index will be used as the starting point. -1 =skew_index0-1 is passed to IQ skew compensation unit 31. After passing through a statistical window, the mean of the current relevant information is compared in unit 84. -1 The absolute value of Mean0. If |Mean -1 If |≤|Mean0|, then skew_index -1 =skew_index0-1 is fed back to compensation module 31, as shown in units 84, 86, and 89, indicating the end of this dynamic tracking, with the tracking adjustment count being 1. Figure 9 IQ skew dynamic tracking diagram (right) showing points A* and B*; if |Mean -1 |>|Mean0|, such as Figure 9 The IQ skew dynamic tracking diagram (right) shows points A and B. Then, using skew_index0 as the baseline, it incrementally increments and tracks another window in reverse. When D=1, skew_index1 = skew_index0 + 1 is fed back to compensation module 31, as shown in units 85, 88, and 89. After one statistical window, in unit 84, the absolute values ​​of the current relevant information mean Mean1 and Mean0 are compared. The tracking adjustment count is 2. Figure 9 The IQ skew dynamic tracking diagram (right) shows points A and C. If |Mean1|≤|Mean0|, then skew_index1=skew_index0+1 is fed back to compensation module 31, as shown in units 84, 86, and 89, which means the dynamic tracking ends; if |Mean1|>|Mean0|, then skew_index0 is fed back to compensation module 31, as shown in units 85, 87, and 89, which means the dynamic tracking ends.

[0099] After the waiting window ends, through reasonable tracking direction prediction, this application can, with a high probability, achieve the effect of ending the tracking adjustment with only one adjustment during the current tracking adjustment process; assuming that the tracking direction prediction is inaccurate, it will scan each of the two directions once. The solution provided by this application can reduce tracking time and dynamically reduce circuit power consumption. The solution provided by this application can fine-tune the IQ skew, gradually approaching the true value, and can track changing IQ skew values ​​with high tracking accuracy.

[0100] The simulation conditions were 16QAM modulation format, baud rate of 62.83 GHz, 1.33x upsampling, sampling factor of 32, with a -1 ps skew impairment added to the signal, a scanning window of 2^20 data beats, and a frequency offset of 1 GHz. The simulation results are as follows. Figure 10 As shown in the dual-axis graph, the left y-axis represents the absolute value of the mean of IQ-related information, and the right y-axis represents the estimated IQ skew value, skew_index. The dynamic tracking result is skew_index = -3, corresponding to a skew value of -1.12 ps. The minimum adjustment precision during estimation is 0.37 ps, and the simulation error for this dynamic scan is 0.12 ps.

[0101] IQskew=skew_index / (1.33×62.83e 9 ×32)×10 12 ≈-1.12ps.

[0102] Example 4

[0103] This embodiment is a specific implementation of scheme (3) in embodiment 1, such as Figure 4 The scheme (3) shown uses a dynamic scanning unit and a dynamic tracking unit to estimate the IQ skew, and consists of an IQ-related information acquisition unit 41, a dynamic scanning unit 42, a dynamic tracking unit 43, and a skew estimation output 45. This scheme can be used as the default scheme for dynamic estimation of IQ skew, and can compromise the needs of embodiments 2 and 3, and can be applied to most scenarios. This embodiment combines the two schemes of embodiments 2 and 3. After completing the dynamic scanning, the software can optionally configure and start the dynamic tracking module. When the software is configured to start dynamic tracking, the final IQ skew estimate output by the dynamic scanning unit 42 is used as the initial estimate skew_index0 of the dynamic tracking unit 43, realizing the transition from rapid estimation of IQ skew by dynamic scanning to long-term estimation by dynamic tracking.

[0104] Based on Example 2, after the dynamic scanning is completed in Example 2, the dynamic scanning process will not be described again. The result is passed to the compensation module 31 for compensation. The software can optionally configure the start of the dynamic tracking module, configure the number of statistical windows for dynamic tracking and the number of data frames used for the statistical windows. The tracking process will not be described again.

[0105] This embodiment first uses a dynamic scanning scheme to quickly obtain the estimated IQ skew value, and then uses a dynamic tracking scheme to fine-tune it, while tracking the changes in IQ skew caused by ambient temperature or other factors.

[0106] The simulation conditions were 16QAM modulation format, baud rate of 62.83 GHz, 1.33x upsampling, sampling factor of 32, with a 2 ps skew impairment added to the signal, a scanning window of 2^20 data beats, and a frequency offset of 1 GHz. The simulation results are as follows. Figure 11 As shown in the dual-axis graph, the left y-axis represents the absolute value of the mean of IQ-related information, and the right y-axis represents the estimated IQ skew value, skew_index. The dynamic tracking result shows skew_index = 5, corresponding to a skew value of 1.87 ps. The minimum adjustment precision during estimation is 0.37 ps, and the simulation error for this dynamic scan is 0.13 ps.

[0107] IQskew=skew_index / (1.33×62.83e 9 ×32)×10 12 ≈1.87ps.

[0108] Example 5

[0109] This embodiment is a specific implementation of scheme (4) in embodiment 1, such as Figure 4 The scheme shown (4) can estimate IQ skew using a dynamic scanning unit and an anomaly monitoring unit; it can estimate IQ skew using a dynamic tracking unit and an anomaly monitoring unit; it can estimate IQ skew using a dynamic scanning unit, a dynamic tracking unit, and an anomaly monitoring unit. This scheme can be used as a supplementary estimation scheme to Embodiments 2, 3, and 4.

[0110] In this embodiment, the IQ-related information acquisition unit 41 obtains the IQ-related information of each statistical window, the unit 122 obtains the absolute value information of the difference between the correlation values ​​of adjacent statistical windows, and the comparison unit 123 compares the absolute value of the current difference with k times the absolute value of the previous difference. If the absolute value of the current difference exceeds k times the absolute value of the previous difference, it indicates that there is an abnormal jump in the IQ skew in the signal at this time, and the unit 124 outputs an abnormal alarm. The software can choose whether to reset and restart the entire dynamic scanning and dynamic tracking process to reduce the abnormal recovery time.

[0111] like Figure 12 As shown, the relevant information receiving unit 120 receives relevant information values. The correlation value mean calculation unit 121 and the correlation value mean calculation unit 51 have the same function, and 120 and 121 can be reused with 50 and 51. At the end of the statistical window, unit 122 obtains the absolute value of the difference between the correlation values ​​of adjacent statistical windows, and compares it with k times the absolute value of the previous difference in unit 123. k is configured by the software according to the usage scenario, and the default value is 8. If the current absolute value is greater than k times the absolute value of the previous difference, it indicates that there is an abnormal jump in the IQ skew in the signal. Unit 124 sets the abnormal alarm to 1. The software can choose whether to reset and restart the entire dynamic scanning and dynamic tracking process to reduce the abnormal recovery time.

[0112] Example 6

[0113] Furthermore, the present invention provides an IQ skew estimation device, including an IQ skew compensation unit, other damage compensation unit, relevant information statistics unit, and IQ skew estimation update unit, wherein the IQ skew estimation update unit estimates the IQ skew using the above-described method.

[0114] The relevant information statistics unit uses the following expression to count relevant value information:

[0115]

[0116] Where n is the sampling time, N is the number of times the relevant information of the sample point is calculated, and K is the weight information.

[0117] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for estimating IQ skewness, characterized in that, Based on the mean of statistically relevant information from the I-channel and Q-channel data, the index value skew_index corresponding to IQ skew is estimated using a dynamic scanning unit. Then, a dynamic tracking unit tracks and adjusts the skew_index based on the IQ skew value. Finally, a skew estimate is output based on the skew_index. The step of calculating the average of relevant information based on the I-path and Q-path data specifically includes: ,in The number of calculations is performed for the relevant information of the sample points. For weight information, It is the nth sampling signal of the in-phase component. It is the sampled signal of the nth pair of orthogonal components; The step of estimating the skew_index value corresponding to the IQ skew value using a dynamic scanning unit specifically includes: determining the initial scanning direction based on the sign of the mean of the initial relevant information, up to the . Second and third If the signs of the secondary relevant information are the same and different from the initial sign, then determine whether the signs of the first and second relevant information means are the same to decide whether a reverse scan is needed at the initial skew value; the scan ends when the signs of the first and second relevant information means are the same, and the second relevant information means are different from the initial sign. Second and third The mean sign of the secondary related information is the same and the sign is opposite to that of the mean sign of the first and second related information. When the scan ends, the skew_index value corresponding to the mean of the related information with the smallest absolute value is taken as the dynamic scan result of the IQ skew value. When the dynamic tracking unit tracks and adjusts the skew_index of the IQ skew value: the direction prediction unit adjusts the skew_index based on the value within the waiting window. The average of the compensated relevant information is used to predict the starting direction for each dynamic tracking iteration; the starting direction is positive, compared with... and The output is the absolute value of the mean of the compensated relevant information. If the current absolute value is small, the dynamic tracking ends. As an estimate; if the current absolute value is large, then compare it with... and The absolute value of the mean of the compensated relevant information is used as the output, with the smaller absolute value corresponding to the IQ skew estimate. The entire tracking process with a negative initial direction is the same as described above, except that the direction is reversed. To dynamically track the initial value of IQ skew.

2. The IQ skew estimation method as described in claim 1, characterized in that, Based on the mean of statistical information of the I-path and Q-path data, the skew_index value corresponding to the IQ skew value is estimated using the dynamic scanning unit, and the skew estimate is output based on the skew_index.

3. The IQ skew estimation method as described in claim 1 or 2, characterized in that, It also includes anomaly detection, specifically: obtaining the absolute value information of the difference between relevant values ​​in adjacent statistical windows. If the absolute value of the current difference exceeds k times the absolute value of the previous difference, it indicates that there is an abnormal jump in IQ skew at this time, and the software chooses whether to reset and restart the scanning and tracking process.

4. The IQ skew estimation method as described in claim 1 or 2, characterized in that, When the interference in the channel is below a preset value, the method for directly calculating the IQ skew value is as follows: in The number of calculations is performed for the relevant information of the sample points. For weight information, For frequency offset information, This is the IQ skew value.

5. An IQ skew estimation device, characterized in that, It includes an IQ skew compensation unit, other damage compensation units, relevant information statistics units, and an IQ skew estimation and update unit, wherein the IQ skew estimation and update unit estimates the IQ skew using the method described in any one of claims 1 to 4.

6. The IQ skew estimation device as described in claim 5, characterized in that, The relevant information statistics unit uses the following expression to calculate the average value of relevant information: in The number of calculations is performed for the relevant information of the sample points. This is weight information.

Images