A method for estimating I / Q time skew based on mirror reference signals

Abstract

The application discloses an I / Q time skew estimation method based on a mirror reference signal and belongs to the technical field of wireless communication baseband signal processing. The method comprises the following steps: performing synchronization processing and frequency offset correction on a received OFDM baseband signal, and obtaining an equivalent frequency domain channel response based on a training field or a pilot field; extracting an equivalent channel phase at an effective frequency point and correcting a phase jump between adjacent effective frequency points; pairing mirror frequency points about zero frequency symmetry, calculating an equivalent mirror phase difference; dividing a mirror same direction sample set and a mirror reverse sample set according to a local reference signal polarity product result corresponding to the mirror frequency points; performing one-dimensional linear fitting on the two types of samples respectively, and calculating an I / Q time skew estimation value according to the difference between a mirror same direction slope and a mirror reverse slope. The application can obtain reliable I / Q time skew estimation results without setting a special calibration sequence.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication baseband signal processing technology, specifically relating to a method for estimating the relative time skew between I-path and Q-path by utilizing existing training fields or pilot fields in orthogonal frequency division multiplexing signals, and through equivalent frequency domain channel response acquisition, phase jump correction, mirror frequency pairing, local reference signal polarity grouping, and linear fitting. Background Technology

[0002] As wireless communication systems evolve towards higher bandwidth and transmission rates, orthogonal frequency division multiplexing (OFDM) technology has been widely applied in broadband communication systems. Ideally, the I and Q paths in a complex baseband signal should maintain a consistent time base. However, due to factors such as non-ideal analog devices, differences in sampling links, and circuit layout, a relative time skew may occur between the two paths. This skew can cause coupling between the original frequency component and its mirrored frequency component, resulting in an equivalent channel phase change, which in turn affects channel estimation, equalization, and subsequent demodulation performance.

[0003] Currently, I / Q time skew is typically estimated using dedicated calibration signals, additional test sequences, or offline calibration. These methods incur additional overhead and struggle to reflect skew changes caused by variations in device state, temperature, and operating bandwidth. Therefore, a solution remains to be found that allows for I / Q time skew estimation using existing training or pilot fields within the communication frame without altering the existing frame structure or inserting additional dedicated calibration sequences. Summary of the Invention

[0004] To address the problem that existing I / Q time skew estimation methods rely on additional calibration signals, dedicated test sequences, or offline calibration processes, making it difficult to complete skew estimation during normal communication, this invention provides an I / Q time skew estimation method based on a mirror reference signal. The mirror reference signal has effective frequency points on both the positive and negative frequency sides that can form mirror frequency pairs, and the local reference signal corresponding to each mirror frequency pair is known at the receiver. This method reuses existing training fields or pilot fields in orthogonal frequency division multiplexing communication frames, eliminating the need for additional dedicated calibration sequences and changes to the existing frame structure.

[0005] To achieve the above objectives, the present invention adopts the following technical solution.

[0006] Specifically, this method is used to estimate the relative time skew between the I and Q paths in an orthogonal frequency division multiplexed baseband signal that includes a mirror-feature reference signal. The receiver performs synchronization processing and frequency offset correction on the received orthogonal frequency division multiplexing baseband signal to determine the time-frequency position of the training field or pilot field in the received signal. Based on the received signal in the reference signal and the corresponding local reference signal, the equivalent frequency domain channel response on each subcarrier is obtained. DC subcarriers, guard subcarriers, and subcarriers not carrying valid reference signals are removed. The phase values ​​of the channel response at valid frequency points are extracted, and the phase transitions between adjacent valid frequency points are corrected to obtain an equivalent channel phase sequence. The mirror frequency points in the equivalent channel phase sequence are paired, and half of the phase difference corresponding to each pair of mirror frequency points is calculated to obtain an equivalent mirror phase difference set. Based on the polarity product of each mirror frequency point with the corresponding local reference signal, the equivalent mirror phase difference set is divided into a mirror-in-the-direction sample set and a mirror-in-the-direction sample set. Univariate linear fitting is performed on the two types of samples to obtain the mirror-in-the-direction slope and the mirror-in-the-direction slope. The I / Q time skew estimate is calculated based on the difference between the mirror-in-the-direction slope and the mirror-in-the-direction slope.

[0007] As a preferred option, the effective frequency point set The pairing of positive and negative effective frequencies with zero-frequency symmetry is called a mirror frequency pair. Calculate the equivalent image phase difference for each pair of image frequency points:

[0008]

[0009] in, Indicates the effective frequency point on the positive frequency side. The corresponding equivalent channel phase, Indicates the image frequency on the negative frequency side. The corresponding equivalent channel phase. Effective frequency points that fail to form mirror frequency pairs on the positive and negative frequency sides are not included in the equivalent mirror phase difference set. The structure of.

[0010] As a preferred approach, the mirror relationship is determined based on the polarity product of the mirror frequency points with the corresponding local reference signals:

[0011]

[0012] when At that time, the corresponding equivalent mirror phase difference is assigned to the mirror-oriented sample set. ;when At that time, the corresponding equivalent mirror phase difference is assigned to the mirror-reverse sample set. .

[0013] As a preferred approach, the mirror-oriented sample sets are respectively... and mirror reverse sample set Establish a univariate linear fitting relationship:

[0014]

[0015] in, Indicates the slope in the same direction as the mirror image. Indicates the reverse slope of the mirror image. and These represent the corresponding univariate linear fitting intercepts. The I / Q time skew estimate is calculated based on the difference between the mirror-image slopes in the same direction and the mirror-image slopes in the opposite direction.

[0016]

[0017] If there are multiple reference OFDM symbols in the reference signal that can be used for I / Q time skew estimation, the processing described in claims 2 to 6 is performed on each reference OFDM symbol to obtain multiple I / Q time skew estimates, and the multiple I / Q time skew estimates are averaged to obtain the final I / Q time skew estimate.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] 1) This invention utilizes existing training fields or pilot fields in communication frames to perform I / Q time skew estimation, without the need to insert additional dedicated calibration sequences or change the existing frame structure.

[0020] 2) This invention extracts the phase change features of two types of image frequency point samples by pairing image frequency points and grouping local reference signals by polarity, and calculates the I / Q time skew estimate by using the difference between the same-direction slope and the opposite-direction slope of the image, thereby reducing the influence of the common linear phase on the estimation result.

[0021] 3) This invention performs linear fitting based on multiple effective frequency points, which can reduce the impact of local noise and single abnormal frequency points on the estimation results. Attached Figure Description

[0022] Figure 1 This is a flowchart of the I / Q time skew estimation method based on mirror reference signal of the present invention;

[0023] Figure 2 This is a schematic diagram of the mirror frequency pairing and equivalent mirror phase difference construction of the present invention;

[0024] Figure 3 This is a schematic diagram of the mirror frequency point sample grouping and linear fitting of the present invention;

[0025] Figure 4 This is a graph showing the sample grouping and linear fitting results of the mirror frequency points when the I / Q time skew setting is 100ps.

[0026] Figure 5 This is a graph showing the absolute estimation error of the present invention under different I / Q time skew settings. Detailed Implementation

[0027] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be understood that the following embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit the scope of protection of the present invention. Example

[0028] This embodiment proposes an I / Q time skew estimation method based on a mirror reference signal. The receiver performs synchronization processing and frequency offset correction on the received orthogonal frequency division multiplexing baseband signal to determine the time-frequency position of reference signals such as training fields or pilot fields in the received signal. Based on the received signal in the reference signal and the corresponding local reference signal, the equivalent frequency domain channel response on each subcarrier is obtained. DC subcarriers, guard subcarriers, and subcarriers not carrying valid reference signals are removed. The phase values ​​of the channel response at valid frequency points are extracted, and the phase transitions between adjacent valid frequency points are corrected to obtain an equivalent channel phase sequence. The mirror frequency points in the equivalent channel phase sequence are paired, and half of the phase difference corresponding to each pair of mirror frequency points is calculated to obtain an equivalent mirror phase difference set. Based on the polarity product of each mirror frequency point with the corresponding local reference signal, the equivalent mirror phase difference set is divided into a mirror-in-direction sample set and a mirror-reverse sample set. Using the corresponding subcarrier physical frequency as the independent variable, univariate linear fitting is performed on the two types of samples to obtain the mirror-in-direction slope and the mirror-reverse slope. The relative time skew between the I-path and Q-path is estimated based on the difference between the mirror-in-direction slope and the mirror-reverse slope.

[0029] Let the ideal complex baseband signal before entering the I / Q branch of the receiver be:

[0030]

[0031] in, Indicates the I-channel signal component. This represents the Q-channel signal component. When there is a time skew between the Q-channel and the I-channel... At that time, the output signal of the receiving end is represented as:

[0032]

[0033] in, This indicates that the Q path is delayed relative to the I path. This indicates that path Q is ahead of path I. Let... and They represent and The frequency domain signal. Due to and Since all signals are real-valued, therefore:

[0034]

[0035] Therefore, we can conclude that:

[0036]

[0037] Combining the time-shift property of Fourier transform: Substitute it into From the frequency domain expression, we get:

[0038]

[0039] when When, the above equation degenerates into ;when At the same time, a mirror frequency component will also appear in the received signal. Therefore, I / Q timing skew can cause coupling between the original frequency component and the mirrored frequency component.

[0040] Let frequency The local reference signal at that location is The channel frequency response when there is no I / Q time skew is The reference signal before passing through the wireless channel and entering the receiver's I / Q branch is represented as:

[0041]

[0042] The corresponding image frequency components are represented as follows:

[0043]

[0044] The receiver obtains the equivalent frequency domain channel response based on the received reference signal and the corresponding local reference signal:

[0045]

[0046] This equation shows that the equivalent frequency domain channel response obtained by conventional channel estimation contains a mirror coupling term caused by I / Q time skew, and therefore the skew information can be extracted by utilizing the phase relationship between the mirror frequency points.

[0047] This embodiment uses the High Efficiency Long Training Field (HE-LTF) from the IEEE 802.11ax (Wi-Fi 6) standard as the reference signal. The HE-LTF is a known sequence used for channel estimation, and its local reference signal at each effective frequency point is a subset of... The bipolar real-valued sign of , therefore its conjugate term satisfies Therefore, the product relationship between any mirror frequency point and its corresponding local reference signal is defined as follows:

[0048]

[0049] when When the mirror frequency points correspond to the two local reference signals with the same polarity; when At this time, the polarities of the two are opposite. Since the product of different polarities will change the sign of the mirror coupling term, the two sample sets need to be processed separately. To illustrate the role of grouping, the linear phase principal term involved in estimating the channel response within the frequency band is expressed as:

[0050]

[0051] in, Indicates the channel amplitude response. This represents a common initial phase that is independent of frequency. This represents the common equivalent time delay caused by propagation delay, timing deviation, and synchronization residual. Within the range of mirror frequencies involved in the estimation, considering the relatively gradual changes in channel amplitude at the positive and negative mirror frequencies, and the fact that the frequency-independent common initial phase can be incorporated into the subsequent fitting intercept, it can be approximated as... When the product of the mirror frequency point and the corresponding local reference signal is At that time, the equivalent frequency domain channel response is approximately expressed as:

[0052]

[0053] Its phase can be expressed as When the product of the mirror frequency point and the corresponding local reference signal is At that time, the equivalent frequency domain channel response is approximately expressed as:

[0054]

[0055] Its phase is Therefore, the mirror-image, same-direction sample set mainly contains the common phase slope, while the mirror-image, opposite-direction sample set, in addition to the common phase slope, also contains the additional phase slope caused by the I / Q time skew. By fitting the two sample sets separately, the influence of the common equivalent time delay can be eliminated by subtracting the slopes.

[0056] The derivation describes the theoretical relationship under continuous frequency variables. In actual processing, the receiver only needs to process the discrete effective frequency points occupied by the reference signal. Let the... A reference signal at the effective frequency point The received signal at the location is The corresponding local reference signal is The equivalent frequency domain channel response is obtained using the least squares algorithm:

[0057]

[0058] in, Indicates the first The i-th subcarrier of OFDM symbol The equivalent frequency domain channel response at that point, Indicates the first The set of reference signal subcarriers for each OFDM symbol. For simplicity, the reference symbol subscripts will be omitted thereafter. The equivalent frequency domain channel response is denoted as .

[0059] After removing DC subcarriers, guard subcarriers, and subcarriers that do not carry valid reference signals, the set of valid frequency points is obtained, arranged from low to high frequency. .in, This represents the total number of effective frequency points involved in the estimation. The phase value of the channel response at each effective frequency point is extracted using the four-quadrant arctangent function.

[0060]

[0061] Because the phase value is limited by its range, adjacent effective frequency points may be close to each other. The non-physical jump. To avoid this jump affecting subsequent slope calculations, the phase difference between adjacent phases is calculated sequentially along the positive direction of the frequency axis. The correction amount is determined according to the following formula:

[0062]

[0063] make The cumulative correction amount is The corrected equivalent channel phase is .

[0064] See Figure 2 During the mirror frequency pairing process, for any effective frequency point on the positive frequency side... Find the effective frequency point on the negative frequency side that is symmetrical about zero frequency. If corresponding valid frequency points exist on both the positive and negative frequency sides, then a mirror frequency pair is formed; if no mirror frequency pair is formed, then the valid frequency point does not participate in the construction of the equivalent mirror phase difference set. The equivalent mirror phase difference is obtained by subtracting the equivalent channel phase corresponding to the mirror frequency pair and dividing by 2.

[0065]

[0066] The equivalent mirror phase difference set is obtained:

[0067]

[0068] See Figure 3 For any sample in the equivalent mirror phase difference set D In this embodiment, the sample is paired with its corresponding mirror frequency point. The mirror relationship is determined by the product of the local reference signal polarity:

[0069]

[0070] The samples are divided into mirror-image and same-direction sample sets. and mirror reverse sample set .in, The samples were classified , The samples were classified .

[0071]

[0072] Will and The samples in the sample list are renumbered as follows:

[0073]

[0074] in, and Representing the mirror image and same direction sample sets respectively The Middle The subcarrier physical frequency and equivalent mirror phase difference corresponding to each sample express The number of samples in the sample; and Representing the mirrored and reversed sample sets respectively The Middle The subcarrier physical frequency and equivalent mirror phase difference corresponding to each sample express The number of samples in the sample.

[0075] In linear fitting, the corresponding subcarrier physical frequency is used as the independent variable. For the mirror-image in-direction sample set... and mirror reverse sample set , uniformly recorded as ,in, ;when hour, Represents the set of mirror-image, same-direction samples; when hour, Let represent the mirror image set. The class sample set contains The nth sample, the th The physical frequency of the negative frequency side subcarrier and the equivalent mirror phase difference corresponding to each sample are respectively: and .

[0076] For the The univariate linear fitting relationship of a set of samples is expressed as:

[0077]

[0078] in, and They represent the first The fitting slope and intercept for each class of samples. Based on the least squares criterion, define the objective function as the sum of squared fitting residuals:

[0079]

[0080] For parameters and Taking the partial derivatives separately and setting them to zero, we get:

[0081]

[0082] in, and They represent the first The mean physical frequency of the subcarrier and the mean equivalent mirror phase difference of the sample are, respectively:

[0083]

[0084] Separately and This yields the univariate linear fitting relationship between the mirror-image and in-direction sample sets and the mirror-image and reverse-direction sample sets:

[0085]

[0086] in, Indicates the slope in the same direction as the mirror image. This represents the reverse slope of the mirror image. This takes into account the presence of residual common delay. and I / Q time skew When the local reference signal polarity multiplication and integration are applied, the equivalent mirror phase difference between the two sample sets exhibits different theoretical phase characteristics. Based on the aforementioned theoretical model derivation, the equivalent mirror phase difference between the two sample sets approximately satisfies:

[0087]

[0088] Therefore, the fitting slopes are approximately as follows: and After subtracting the slopes from the two sets, we have Thus, the I / Q time skew estimate is obtained:

[0089]

[0090] By calculating the difference between the mirror-image in-direction slope and the mirror-image reverse slope, the linear phase slope corresponding to the common delay can be canceled out, retaining only the additional phase slope caused by the I / Q time skew. When there are multiple reference OFDM symbols in the reference signal that can be used for I / Q time skew estimation, the processing described in claims 2 to 6 is performed on each reference OFDM symbol to obtain multiple I / Q time skew estimates. These multiple I / Q time skew estimates are then averaged to obtain the final I / Q time skew estimate. If the output is in picoseconds, then... . Example

[0091] This embodiment verifies the I / Q time skew estimation method based on the mirror reference signal. This embodiment uses only the IEEE 802.11ax High Efficiency Long Training Field (HE-LTF) as the reference signal for illustration and does not constitute a limitation on the reference signal type, communication standard, or system bandwidth of this invention. Specifically, the IEEE 802.11ax High Efficiency Single User Physical Layer Protocol Data Unit is used, with a transmission bandwidth set to 40 MHz. After the baseband signal is generated, the transmitting end sets a time skew for the Q path relative to the I path, and the receiving end completes the estimation according to the method described in Embodiment 1.

[0092] Number of Fast Fourier Transform points at a 40MHz bandwidth for The physical subcarrier spacing is:

[0093]

[0094] See Figure 4 , Figure 4 The image frequency sample grouping and linear fitting results were set when the I / Q time skew was 100 ps. Specifically, as follows: Figure 4 As shown in (a), its corresponding mirror image set is the same direction sample set. The fitting results; such as Figure 4 As shown in (b) above, its corresponding mirror image set The fitting results are as follows: the horizontal axis of both sub-graphs is the physical frequency of the subcarrier, and the vertical axis is the equivalent mirror phase difference.

[0095] As can be seen from the figure, the set of mirror-image, same-direction samples The fitted slope is Approaching zero; mirrored reverse sample set The fitted slope is The equivalent mirror phase difference exhibits a significant linear variation with the physical frequency of the subcarrier. Based on the difference between the mirror in-direction slope and the mirror in-reverse slope, the estimated I / Q time skew value is calculated to be 99.40 ps, ​​with an absolute error of 0.60 ps compared to the set value of 100 ps.

[0096] To verify the estimation effect of the method under different time skew conditions, the I / Q time skew setting value was further set to... The absolute estimation error under different setpoint conditions is as follows: Figure 5 As shown. In the tested to Within the range, the maximum absolute estimation error is .

[0097] The above results show that, under the parameter configuration adopted in this embodiment, the method can reuse the existing HE-LTF in the communication frame to complete the I / Q time skew estimation without the need to insert an additional dedicated calibration sequence; through mirror frequency pairing, polarity grouping and group linear fitting, a time skew estimation result close to the set value can be obtained.

[0098] Scope of application and transformation methods

[0099] This invention is not limited to specific communication standards, specific system bandwidths, specific modulation schemes, or specific reference signal structures. Any orthogonal frequency division multiplexing system that can obtain the frequency domain reference response, construct positive and negative frequency mirror pair, and group the equivalent mirror phase difference based on the product of the polarities of the local reference signals corresponding to the mirror pair, can use this invention for I / Q time skew estimation.

[0100] In other implementations, the effective frequency range involved in the estimation can be adjusted according to the actual subcarrier configuration, the construction method of the mirror frequency pairs can be adjusted according to the distribution of effective frequency points, and the construction method of the mirror in-direction sample set and the mirror in-reverse sample set can be adjusted according to the known polarity relationship of the reference signal.

[0101] In other embodiments, the reference signal used to construct the equivalent frequency domain channel response can be a training field, a pilot field, a reference sequence, or other frequency domain symbols with known mirror relationships; the reference frequency points involved in the estimation can be all effective frequency points or some effective frequency points that meet preset conditions.

[0102] The adjustments made in the above embodiments do not change the technical solution of estimating I / Q time skew by using the difference between the mirror-image slope in the same direction and the mirror-image slope in the opposite direction, and therefore should all fall within the protection scope of this invention.

Claims

1. A method for estimating I / Q time skew based on a mirror reference signal, used to estimate the relative time skew between the I and Q paths in an orthogonal frequency division multiplexing (OFDM) baseband signal containing a mirror feature reference signal, characterized in that, The steps include: S1. Synchronizing and frequency offset correcting the received orthogonal frequency division multiplexing baseband signal to determine the time-frequency position of the reference signal in the received signal, wherein the reference signal includes a training field or a pilot field; S2. Obtain the equivalent frequency domain channel response on each subcarrier based on the received signal in the reference signal and the corresponding local reference signal; S3. Using the equivalent frequency domain channel response as input, remove DC subcarriers, guard subcarriers, and subcarriers that do not carry effective reference signals, extract the phase value of the channel response at the effective frequency point where the reference signal is located, and correct the phase jump between adjacent effective frequency points to obtain the equivalent channel phase sequence. S4. Pair up the mirror frequency points in the equivalent channel phase sequence, calculate half of the phase difference corresponding to each pair of mirror frequency points, and obtain the equivalent mirror phase difference set; S5. Based on the polarity product of each mirror frequency point with the corresponding local reference signal, divide the equivalent mirror phase difference set into a mirror in-direction sample set and a mirror in-reverse sample set; S6. Using the corresponding subcarrier physical frequency as the independent variable, perform univariate linear fitting on the mirror in-direction sample set and the mirror in-reverse sample set to obtain the mirror in-direction slope and the mirror in-reverse slope; S7. Calculate the I / Q time skew estimate based on the difference between the mirror in-direction slope and the mirror in-reverse slope.

2. The method according to claim 1, characterized in that, Step S2 specifically includes: the equivalent frequency domain channel response is calculated using a conventional channel estimation algorithm; specifically, a least squares channel estimation algorithm can be used, the expression of which is: in, Indicates the first The i-th subcarrier of OFDM symbol The equivalent frequency domain channel response at that point, Indicates the first OFDM symbols at frequency The reference signal received at the location, Indicates the first OFDM symbols at frequency The local reference signal at that location, Indicates the first A set of reference signal subcarriers for OFDM symbols.

3. The method according to claim 1, characterized in that, Step S3 specifically includes: using the output of step S2... Equivalent frequency domain channel response of OFDM symbols For input (for simplicity, in the selected number) After one OFDM symbol, the subsequent symbol subscripts are omitted. After removing DC subcarriers, guard subcarriers, and subcarriers that do not carry valid reference signals, the set of valid frequency points arranged from low to high frequency is obtained. ,in, This represents the total number of effective frequency points involved in the estimation; the phase value of the channel response at each effective frequency point is extracted using the four-quadrant arctangent function. in, This represents the operator for finding the imaginary part of a complex number. This represents the operator for finding the real part of a complex number; it calculates the phase difference between adjacent numbers sequentially along the positive direction of the frequency axis. The correction amount is determined according to the following formula: make The cumulative correction amount is The corrected equivalent channel phase is The equivalent channel phase sequence for channel estimation is thus output: .

4. The method according to claim 1, characterized in that, Step S4 specifically includes: using the equivalent channel phase sequence output in step S3. For input; set of effective frequency points The pairing of positive and negative effective frequencies with zero-frequency symmetry is called a mirror frequency pair. Calculate the equivalent image phase difference for each pair of image frequency points: ,in, Indicates the effective frequency point on the positive frequency side. The corresponding equivalent channel phase, Indicates the image frequency on the negative frequency side. The corresponding equivalent channel phase; thus, the equivalent image phase difference set is obtained: ,in, This represents the set of effective frequency points on the positive frequency side that have successfully achieved mirror pairing. This indicates the physical frequency of the corresponding negative frequency side of the image frequency point. Effective frequencies that fail to form a mirror frequency pair on the positive and negative frequency sides do not participate in the equivalent mirror phase difference set. The structure of.

5. The method according to claim 1, characterized in that, Step S5 specifically includes: for the equivalent mirror phase difference set any sample in Based on the mirror frequency point corresponding to the sample The mirror relationship is determined by the product of the local reference signal polarity: in, Indicates the effective frequency point on the positive frequency side. The corresponding local reference signal, Indicates the image frequency on the negative frequency side. The corresponding local reference signal; when At that time, the sample Included in the mirror image and same direction sample set ;when At that time, the sample Included in the mirror reverse sample set ; Will and The samples in the sample list are renumbered as follows: in, and Representing the mirror image and same direction sample sets respectively The Middle The subcarrier physical frequency and equivalent mirror phase difference corresponding to each sample express The number of samples in the sample; and Representing the mirrored and reversed sample sets respectively The Middle The subcarrier physical frequency and equivalent mirror phase difference corresponding to each sample express The number of samples in the set; subcarriers that fail to form a valid mirror frequency pair or whose corresponding local reference signals are invalid are not included in the mirror in-direction sample set. and mirror reverse sample set The structure of.

6. The method according to claim 1, characterized in that, Step S6 specifically includes: using the mirror-oriented sample set output in step S5. and the mirror reverse sample set As input, perform univariate linear fitting on both types of samples respectively; for the set of mirror-image, same-direction samples... Calculate the average physical frequency of its subcarriers. Mean phase difference with equivalent mirror image : Based on the least squares criterion, the mirror-image slopes are obtained through fitting. Same-direction intercept as mirror : Similarly, for the aforementioned mirrored reverse sample set Calculate the average physical frequency of its subcarriers. Mean phase difference with equivalent mirror image : Based on the least squares criterion, the mirror-image reverse slope is obtained by fitting. and mirror reverse intercept : The univariate linear fitting relationship is obtained: in, and These represent the fitted equivalent mirror phase differences corresponding to the two sample sets, respectively. This indicates the physical frequency of the subcarrier.

7. The method according to claim 1, characterized in that, Step S7 specifically includes: based on the mirror-image slope. and mirror reverse slope Calculate the I / Q time skew estimate: in, This represents the I / Q time skew estimate, in seconds. When multiple reference OFDM symbols exist in the reference signal that can be used for I / Q time skew estimation, the corresponding I / Q time skew estimate is obtained based on each reference OFDM symbol, and the multiple I / Q time skew estimates are averaged to obtain the final I / Q time skew estimate. If the output is in picoseconds, then... .

Images