A high-precision timestamp detection device and method for wireless time synchronization

Abstract

This invention discloses a high-precision timestamp detection device for wireless time synchronization, comprising a transmitting baseband processing module, a transmitting timestamp module, and a transmitting radio frequency module. The receiving end includes a receiving radio frequency module, a frame decision module, a receiving timestamp recording module, a receiving timestamp correction module, and a receiving baseband processing module. It also discloses a high-precision timestamp detection method for wireless time synchronization. This method processes the received signal using correlation algorithms to obtain a reference time corresponding to an integer multiple of the signal arrival time of the sampling period delay. Subsequently, a Lorentz distribution function is used to fit the acquired discrete baseband signal to obtain a sub-sampling period delay, and the reference time is further corrected, thereby effectively improving the accuracy of the received timestamp. This method effectively overcomes the limitation of sampling rate in traditional timestamp detection, improves the resolution of the received timestamp, and significantly enhances the accuracy of the wireless time synchronization system.

Description

Technical Field

[0001] This invention relates to the field of time synchronization technology, specifically to a high-precision timestamp detection device for wireless time synchronization, and also to a high-precision timestamp detection method for wireless time synchronization. Background Technology

[0002] In wireless time synchronization, the timestamp serves as the core synchronization information for calculating clock skew between nodes, and its accuracy has a decisive impact on the overall system synchronization precision. Comparatively, obtaining the transmission timestamp is relatively simple. Since the transmission time of synchronization information is constrained by the channel access mechanism, and the physical layer processing delay is usually relatively fixed, a highly accurate transmission timestamp can be obtained simply by recording the local clock when the physical layer begins processing the synchronization information. In contrast, the accuracy of the reception timestamp mainly depends on the synchronization precision of the frame synchronization. Limited by factors such as low signal-to-noise ratio environments and frequency offsets, the estimation performance of frame arrival times is easily interfered with, thus affecting the accuracy of the reception timestamp. In wireless communication-based time synchronization systems, the process of obtaining the timestamp at the receiver mainly relies on detecting the maximum value of the correlation peak in the physical layer data frame synchronization sequence.

[0003] Existing correlation-based time synchronization methods can be divided into two categories: autocorrelation and cross-correlation. Autocorrelation methods utilize the inherent characteristics of the received signal to determine the time synchronization position, while cross-correlation methods generate a reference sequence identical to the transmitted sequence locally, perform correlation operations with the received signal, extract the correlation peak, and complete physical layer time synchronization through threshold detection. However, correlation operations are typically performed in the digital domain after signal sampling, thus being limited by the sampling clock; that is, the resolution of the received timestamp cannot be better than one sampling clock cycle, thereby limiting the accuracy of time synchronization.

[0004] Therefore, this invention proposes a high-precision timestamp detection method for wireless time synchronization. Based on a high-precision clock source, after completing the relevant calculations, the sampled discrete signal is further subjected to Lorentz fitting, and then the fitted signal is differentiated to achieve the estimation of the fractional sampling period delay of the frame arrival time, effectively improving the accuracy of the received timestamp and thus improving the overall performance of the wireless time synchronization system. Summary of the Invention

[0005] This invention belongs to the field of time and frequency technology, and provides a high-precision timestamp detection device for wireless time synchronization, as well as a high-precision timestamp detection method for wireless time synchronization.

[0006] The above-mentioned objective of the present invention is achieved through the following technical solution:

[0007] A high-precision timestamp detection device for wireless time synchronization includes a transmitter and a receiver. The transmitter's baseband processing module converts the message into a baseband signal and inputs it to the transmitter's timestamp module and the transmitter's radio frequency module. The transmitter's timestamp module detects and records the arrival time of the transmitter's baseband signal as the transmission time t1. The transmitter's radio frequency module converts the transmitter's baseband signal into a radio frequency signal for transmission.

[0008] The radio frequency (RF) signal is received by the receiver's RF module and converted into a baseband discrete signal before being sent to the frame decision module. The frame decision module performs correlation calculations to obtain the synchronization flag signal (sync_flag) corresponding to the baseband discrete signal. It then inputs the baseband discrete signal, the synchronization flag signal (sync_flag), and the corresponding received signal symbol bit to the receiver's timestamp recording module. The receiver's timestamp recording module, in conjunction with the synchronization flag signal (sync_flag), performs frame detection on the baseband discrete signal to obtain the reference time T for the arrival of the corresponding baseband discrete signal in the message. rx_record and reference time T rx_record and the baseband discrete signal at reference time T rx_record The corresponding sampling point data is input to the receiver's timestamp correction module; the receiver's timestamp correction module uses the data from the baseband discrete signal and the reference time T... rx_record The subsampling period delay is calculated based on the corresponding sampling point data, and then compared with the reference time T. rx_record Make corrections to obtain the corrected arrival time T. rx_corrected Meanwhile, the baseband discrete signal is input to the receiving baseband processing module via the frame decision module for demodulation.

[0009] A high-precision timestamp detection method for wireless time synchronization, utilizing the high-precision timestamp detection device for wireless time synchronization as described above, is characterized by comprising the following steps:

[0010] Step 1: The baseband processing module of the transmitting end sends the baseband signal of the transmitting end corresponding to the message to the timestamp module and the radio frequency module of the transmitting end; the timestamp module of the transmitting end detects and records the arrival time of the baseband signal of the transmitting end as the transmission time t1; the radio frequency module of the transmitting end converts the baseband signal of the transmitting end into a radio frequency signal and then transmits it.

[0011] Step 2: The receiver RF module receives the RF signal, converts the RF signal into the receiver's baseband discrete signal and obtains the corresponding received signal symbol bit of the baseband discrete signal. The decision module obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal through correlation calculation.

[0012] The frame decision module inputs the baseband discrete signal, the synchronization flag signal sync_flag, and the received signal symbol bit to the receiving end timestamp recording module;

[0013] The receiver's timestamp recording module combines the synchronization flag signal sync_flag and the message's frame header data pattern to perform frame detection on the baseband discrete signal, obtaining the reference time T for the arrival of the corresponding baseband discrete signal in the message. rx_record and the baseband discrete signal at reference time T rx_record The corresponding sampled data sequence, in the baseband discrete signal, is related to the reference time T. rx_record The corresponding sampling point data sequence is denoted as the subsampling period delay calculation interval;

[0014] Step 3: The receiving end timestamp correction module calculates the subsampling period delay Δt based on the sampling point data in the subsampling period delay calculation interval;

[0015] Step 4: The receiver timestamp correction module adjusts the reference time T according to the subsampling period delay Δt. rx_record Make corrections to obtain the corrected arrival time T. rx_corrected .

[0016] Obtain the reference time T as described above rx_record Includes the following steps:

[0017] During the inflow of baseband discrete signals, the receiving end timestamp recording module uses a state machine to monitor the entire frame data sequence of the frame header in real time. When the complete frame data sequence of the frame header is detected, the receiving end timestamp recording module pulls up the Flag_demod signal and maintains it for two symbol periods.

[0018] During the period when the Flag_demod signal is high, the receiving end timestamp recording module detects changes in the received signal symbol bit and uses the timestamp of the first signal edge of the received signal symbol bit during the period when the Flag_demod signal is high as the reference time T for data arrival. rx_record Reference time T rx_record The data sequence of sampled points in the baseband discrete signal during the period when the Flag_demod signal is high is input to the receiving end timestamp correction module. The data sequence of sampled points in the baseband discrete signal during the period when the Flag_demod signal is high is the data sequence of sampled points in the baseband discrete signal corresponding to the reference time T. rx_record The corresponding sampling point data sequence.

[0019] As described above, step 3 includes the following steps:

[0020] Step 3.1: The receiving end timestamp correction module searches for the sampling point data within the subsampling period delay calculation interval, finds the maximum amplitude of the sampling point, and records the sampling point position corresponding to the maximum amplitude as the maximum discrete signal position x. pk ;

[0021] Step 3.2: Fit the sampling point data in the subsampling period delay calculation interval based on the Lorentz distribution function to obtain the peak position x0 of the subsampling period delay calculation interval;

[0022] Step 3.3: Calculate the peak position x0 relative to the position of the maximum discrete signal using the following formula. pk Offset δ:

[0023] δ=x0-x pk ,

[0024] When the baseband sampling clock frequency is f s Then the subsampling period delay Δt is:

[0025]

[0026] The corrected arrival time, as described above, is obtained based on the following formula:

[0027] T rx_corrected =T rx_record +Δt,

[0028] T rx_corrected This indicates the corrected arrival time.

[0029] As described above, step 3.2 specifically includes the following steps:

[0030] Substituting the sampling point positions and corresponding sampling point amplitudes of the sampling point data within the subsampling period delay calculation interval into the following probability density function model, and fitting the model, we obtain the corresponding fitted curve:

[0031]

[0032] Where A is the amplitude coefficient of the distribution, which is the parameter to be fitted; x0 is the peak position, which is the value to be confirmed; γ is the scale parameter, which corresponds to the half-width at half the maximum value of the distribution; x is the sampling point position; and f(x) is the signal amplitude distribution function.

[0033] Taking the derivative of the signal amplitude distribution function f(x), we obtain its first derivative expression:

[0034]

[0035] The value of x obtained by solving f′(x)=0 is the peak position x0.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] This invention proposes a high-precision timestamp detection method for wireless time synchronization, aiming to improve the accuracy of the receiver's timestamp and thus enhance overall time synchronization performance. The method first uses a high-precision frequency source and employs autocorrelation or cross-correlation principles, combined with the correlation calculation output, to achieve a rough estimate of the signal arrival time as an integer multiple of the sampling period delay. Based on this, a Lorentz fitting model is introduced to fit the baseband discrete signal after the correlation calculation, and then the fitted signal is differentiated to obtain the sub-sampling period delay. By correcting the received timestamp at the sub-sampling level, the impact of sampling rate limitations on synchronization accuracy is effectively overcome, providing crucial support for building a high-precision, low-cost wireless time synchronization system.

[0038] This invention employs Lorentz fitting of the waveform and a method of differentiating the fitted signal to obtain more accurate peak values, better fitting results, and higher precision.

[0039] The Lorentz fitting and differentiation method uses the analytical derivative of the fitting function to find the extreme points, which can accurately locate the peak value, and is especially suitable for sharp and symmetrical peak shapes.

[0040] Using multiple points, the signal curve is globally smoothed, effectively filtering out local random noise;

[0041] By relying on function fitting, peak values ​​can still be estimated even with data that is not evenly spaced or has a low sampling rate. This avoids fitting failures or offsets caused by the fixed point spacing affecting the results. Attached Figure Description

[0042] Figure 1 This is a diagram of a wireless time synchronization device.

[0043] Among them, 1-transmit baseband processing module, 2-transmitter timestamp module, 3-transmitter radio frequency module, 4-receiver radio frequency module, 5-frame decision module, 6-receiver timestamp recording module, 7-receiver timestamp correction module, and 8-receive baseband processing module;

[0044] Figure 2 This is a timing diagram of the receiver timestamp recording module algorithm; where S0-S32 represent the 32 bits of the frame header, corresponding to 33 states of the state machine; filtered_sig refers to the received signal symbol bit of any received signal; Timestamp_demod1 represents the arrival timestamp of the previous frame; Rtc_Time represents the arrival timestamp of the current frame, corresponding to the obtained reference time T. rx_record ;

[0045] Figure 3 A magnified view of a portion of the algorithm for the timestamp recording module at the receiving end;

[0046] Figure 4 Lorentz distribution fitting plot for the timestamp correction module at the receiving end;

[0047] Figure 5 The derivative function plot for fitting the Lorentz distribution of the receiver timestamp correction module. Detailed Implementation

[0048] To facilitate understanding and implementation of the present invention by those skilled in the art, the present invention will be further described in detail below with reference to embodiments. It should be understood that the embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0049] Example 1

[0050] A high-precision timestamp detection device for wireless time synchronization, such as Figure 1 As shown, it includes a transmitter and a receiver. The transmitter includes a baseband processing module 1, a timestamp module 2, and a radio frequency module 3. The receiver includes a radio frequency module 4, a frame decision module 5, a timestamp recording module 6, a timestamp correction module 7, and a baseband processing module 8. All of the above modules are implemented by FPGA.

[0051] The baseband processing module 1 at the transmitting end converts the message into a baseband signal at the transmitting end and inputs it to the timestamp module 2 and the radio frequency module 3 at the transmitting end. In this embodiment, the baseband signal at the transmitting end is first input to the timestamp module 2 at the transmitting end. The timestamp module 2 at the transmitting end detects and records the arrival time of the baseband signal at the transmitting end as the transmission time t1. Then, the baseband signal at the transmitting end is input to the radio frequency module 3 at the transmitting end. The radio frequency module 3 at the transmitting end converts the baseband signal at the transmitting end into a radio frequency signal for transmission. Alternatively, the baseband signal at the transmitting end can be input to both the timestamp module 2 and the radio frequency module 3 at the transmitting end simultaneously.

[0052] The receiver's RF module 4 receives the RF signal and converts it into a baseband discrete signal, which is then sent to the frame decision module 5. The baseband discrete signal is a discrete sequence of sampled data points. The frame decision module 5 obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal as the optimal decision point through correlation operations. It then inputs the baseband discrete signal, the synchronization flag signal sync_flag, and the corresponding received signal symbol bit to the receiver's timestamp recording module 6. The receiver's timestamp recording module 6 performs frame detection on the baseband discrete signal in conjunction with the synchronization flag signal sync_flag, thereby detecting and recording the reference time T at which the baseband discrete signal corresponding to the message arrives. rx_record and reference time Trx_record and the baseband discrete signal at reference time T rx_record The corresponding sampling point data is input to the receiving end timestamp correction module 7; the receiving end timestamp correction module 7 uses the baseband discrete signal and the reference time T rx_record The corresponding sampling point data is used to estimate the subsampling period delay, and the reference time T is then calculated based on the subsampling period delay. rx_record Make corrections to obtain the corrected arrival time T. rx_corrected Meanwhile, the baseband discrete signal is input to the receiving baseband processing module 8 through the frame decision module 5, and the receiving baseband processing module 8 demodulates the baseband discrete signal.

[0053] In this embodiment, the message carries timestamp data, and the corrected arrival time T rx_corrected The data is sent to the receiving baseband processing module 8, which will then correct the arrival time T. rx_corrected The signal is embedded in the baseband discrete signal and then demodulated to read the corrected arrival time T. rx_corrected The corresponding timestamp.

[0054] Example 2

[0055] A high-precision timestamp detection method for wireless time synchronization, utilizing the high-precision timestamp detection device for wireless time synchronization described in Example 1, mainly includes two parts: integer multiple sampling period delay estimation and sub-sampling period delay estimation. The integer multiple sampling period delay estimation is used to initially determine the reference time T of the arrival of the baseband discrete signal at the receiving end of the corresponding signal message. rx_record The estimation of subsampling period delay is used to further accurately locate the true arrival time of the signal, thereby achieving subsampling accuracy timestamp correction.

[0056] It should be noted that the present invention applies to both autocorrelation and cross-correlation synchronization methods. While the two methods differ slightly in how they obtain integer multiples of the delay, they are consistent in how they estimate the delay over fractional sampling periods. For ease of explanation, this embodiment uses the autocorrelation method at the receiving end as an example for detailed description.

[0057] A high-precision timestamp detection method for wireless time synchronization includes the following steps:

[0058] Step 1: Check the sending timestamp:

[0059] The transmitting baseband processing module 1 of the transmitting end sends the baseband signal corresponding to the message to the transmitting timestamp module 2 and the transmitting radio frequency module 3; the transmitting timestamp module 2 detects and records the arrival time of the baseband signal of the transmitting end as the transmission time t1; the transmitting radio frequency module 3 converts the baseband signal of the transmitting end into a radio frequency signal and then transmits it; in this embodiment, the message includes a frame header, edge detection, type, number of synchronization cycles, synchronization message, synchronization message verification, timestamp data, first timestamp data verification, and second timestamp data verification, wherein the type is used to indicate which frame message it is, where 00 represents the first frame message, 01 represents the second frame message, 10 represents the third frame message, and 11 represents the fourth frame message; edge detection is used to detect the integer cycle time when the signal of the receiving end arrives;

[0060] Step 2: Detect the sampling period as an integer multiple of the timestamp at the receiving end:

[0061] The receiving end RF module 4 receives the RF signal, converts the RF signal into the receiving end's baseband discrete signal, and obtains the corresponding received signal symbol bit. The decision module 5 obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal through correlation operations. In this embodiment, the frame decision module 5 obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal through correlation operations. The frame decision module 5 inputs the baseband discrete signal, the synchronization flag signal sync_flag, and the received signal symbol bit to the receiving end timestamp recording module 6.

[0062] In the receiver timestamp recording module 6, the core issue is how to accurately record the arrival time of valid data (i.e., the baseband discrete signal corresponding to the message). To ensure the validity of the timestamp of the arrival time of the baseband discrete signal corresponding to the message, the receiver timestamp recording module 6 combines the synchronization flag signal sync_flag and the frame header data mode of the message to perform frame detection on the baseband discrete signal and obtain the reference time Trx_record for the arrival of valid data at the receiver. The specific process is as follows:

[0063] During the inflow of baseband discrete signals, the receiving end timestamp recording module 6 uses a state machine to monitor the entire frame data sequence of the frame header in real time. When a complete frame data sequence of the frame header is detected, the receiving end timestamp recording module 6 pulls the Flag_demod signal high and maintains it for two symbol periods (the symbol period is determined by the data rate and is also related to the modulation and demodulation methods of the transmitting baseband processing module 1 and the receiving baseband processing module 8). While the Flag_demod signal is high, the receiving end timestamp recording module 6 detects changes in the received signal symbol bits (in this embodiment, it detects changes in the received signal symbol bits of the in-phase branch signal filtered_I or the quadrature branch signal filtered_Q), and timestamps the first signal edge of the received signal symbol bits during the period when the Flag_demod signal is high (in this embodiment, the rising edge or falling edge is selected according to the edge detection setting mode set in the message; in this embodiment, the rising edge is used, such as...). Figure 2 As shown, when state s12 is detected, the timestamp of the first rising edge of the received signal symbol bit is recorded as the reference time T for data arrival. rx_record The present invention uses the signal edge of the received signal symbol bit as a reference signal to obtain the reference time T. rx_record The process no longer relies on traditional synchronization flag signals, effectively avoiding errors and uncertainties caused by synchronization delays, thereby further improving the accuracy of time synchronization;

[0064] The receiving end timestamp recording module 6 will reference time T rx_record The sampling point data sequence of the baseband discrete signal during the period when the Flag_demod signal is high is input to the receiving end timestamp correction module 7. The sampling point data sequence of the baseband discrete signal during the period when the Flag_demod signal is high is used as the data sequence of the baseband discrete signal relative to the reference time T. rx_record The corresponding sampling point data sequence is denoted as the subsampling period delay calculation interval.

[0065] Since the received signal symbol bit corresponds to a sampling point of the baseband discrete signal, the reference time T rx_record The accuracy can only be an integer multiple of the sampling period.

[0066] Step 3: Detect the subsampling period delay of the receiver's timestamp:

[0067] In receiver timestamp detection, to overcome sampling rate limitations and improve timestamp accuracy, it is necessary to estimate the subsampling period delay between sampling points. This subsampling period delay reflects the difference between the actual arrival time of the signal and the reference time T. rx_recordThe deviation between these parameters is crucial for achieving subsampling precision time synchronization. Specifically, when the Flag_demod signal, which serves as a synchronization flag, is detected to be high, it indicates that a baseband discrete signal corresponding to a message has arrived. At this time, the receiver's timestamp correction module 7 immediately calculates the subsampling period delay based on the sampled data within the subsampling period delay calculation interval. The specific process is as follows:

[0068] Step 3.1: The receiving end timestamp correction module 7 calculates the sampling point data (such as...) within the subsampling period delay interval. Figure 3 As shown, a search is performed (the sampling point data includes the sampling point position and the sampling point amplitude) to find the maximum value of the sampling point amplitude, and the sampling point position corresponding to the maximum value of the sampling point amplitude is recorded as the maximum discrete signal position x. pk The location of the sampling point is the sampling point number.

[0069] Step 3.2: In order to more accurately estimate the true location of the maximum amplitude value in the subsampling period delay calculation interval and obtain the time estimate beyond the sampling period accuracy, the sampling point data in the subsampling period delay calculation interval is fitted based on the Lorentzian distribution function to obtain the peak position x0 of the subsampling period delay calculation interval.

[0070] Substituting the sampling point positions and corresponding sampling point amplitudes of the sampling point data within the subsampling period delay calculation interval into the following probability density function model, and performing fitting, yields the corresponding fitting curve, as shown in the figure. Figure 4 As shown, the curve exhibits a typical single-peak structure, which can effectively describe the morphological changes of the signal near the peak.

[0071]

[0072] Where A is the amplitude coefficient of the distribution, which is the parameter to be fitted; x0 is the peak position, which is the value to be confirmed; γ is the scale parameter, which corresponds to the half width at half the maximum value of the distribution; x is the sampling point position; f(x) is the signal amplitude distribution function; when fitting, the sampling point position of the sampling point data in the subsampling period delay calculation interval is substituted into x, and the sampling point amplitude is substituted into f(x; x0; γ).

[0073] To further pinpoint the peak position x0, the signal amplitude distribution function f(x) can be differentiated to obtain its first derivative expression:

[0074]

[0075] The derivative function described above reflects the change in the slope of the fitted curve, exhibits odd function characteristics, and takes the value of zero at x = x0. Therefore, the x value obtained by solving for f′(x) = 0 is the peak position x0. The derivative graph is shown below. Figure 5As shown, its zero-point position clearly reflects the location of the signal peak, thus enabling high-precision estimation of the arrival time.

[0076] Step 3.3: Calculate the peak position x0 relative to the position of the maximum discrete signal, which is limited by an integer sampling period, using the following formula. pk Offset δ:

[0077] δ=x0-x pk ,

[0078] When the baseband sampling clock frequency is f s Then the subsampling period delay Δt can be obtained:

[0079]

[0080] Step 4: The receiving end timestamp correction module 7 adjusts the reference time T according to the subsampling period delay Δt. rx_record Make corrections:

[0081] If δ > 0, it means that the locally recorded data has arrived at the reference time T. rx_record If the arrival time is earlier than the actual signal arrival time, it should be corrected backward:

[0082] T rx_corrected =T rx_record +Δt,

[0083] T rx_corrected This indicates the corrected arrival time.

[0084] If δ < 0, it means that the timestamp of the locally recorded data lags behind the actual arrival time of the signal and needs to be corrected backward.

[0085] T rx_corrected =T rx_record -(-Δt).

[0086] Using the above methods, the received timestamp can be used to compensate and correct the fractional sampling period, thereby repairing the time synchronization error introduced by the limited accuracy of the sampling clock, and thus realizing the time synchronization error within one sampling period.

[0087] This invention effectively overcomes the technical bottleneck of traditional correlation detection, which is limited by the sampling rate and struggles to achieve sub-sampling period accuracy, by using a curve fitting method. Furthermore, this method is applicable to both autocorrelation and cross-correlation synchronization methods. While there are some differences in obtaining integer multiples of the sampling period delay, the estimation method for the sub-sampling period delay remains consistent.

[0088] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A high-precision timestamp detection method for wireless time synchronization, characterized in that, Includes the following steps: Step 1: The baseband processing module (1) of the transmitting end sends the baseband signal of the transmitting end corresponding to the message to the timestamp module (2) and the radio frequency module (3) of the transmitting end; the timestamp module (2) of the transmitting end detects and records the arrival time of the baseband signal of the transmitting end as the transmission time t1; the radio frequency module (3) of the transmitting end converts the baseband signal of the transmitting end into a radio frequency signal and then transmits it; Step 2: The receiver RF module (4) receives the RF signal, converts the RF signal into the baseband discrete signal of the receiver and obtains the received signal symbol bit corresponding to the baseband discrete signal. The decision module (5) obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal through correlation operation. The frame decision module (5) inputs the baseband discrete signal, the synchronization flag signal sync_flag and the received signal symbol bit to the receiving end timestamp recording module (6). The receiver timestamp recording module (6) combines the synchronization flag signal sync_flag and the frame header data mode of the message to perform frame detection on the baseband discrete signal and obtain the reference time of arrival of the corresponding baseband discrete signal of the message. and the baseband discrete signal at the reference time The corresponding sampled data sequence, in the baseband discrete signal and the reference time The corresponding sampling point data sequence is denoted as the subsampling period delay calculation interval; Step 3: The receiving end timestamp correction module (7) calculates the subsampling period delay based on the sampling point data in the subsampling period delay calculation interval. ; Step 4, Receiver timestamp correction module (7) adjusts the time according to the subsampling period delay. Reference time Make corrections to obtain the corrected arrival time. ; Step 3 includes the following steps: Step 3.1: The receiving end timestamp correction module (7) searches for the sampling point data in the subsampling period delay calculation interval, finds the maximum amplitude of the sampling point, and records the sampling point position corresponding to the maximum amplitude of the sampling point as the position of the maximum discrete signal. ; Step 3.2: Fit the sampling point data in the subsampling period delay calculation interval based on the Lorentz distribution function to obtain the peak position of the subsampling period delay calculation interval. ; Step 3.3: Calculate the peak position using the following formula. Relative to the position of the maximum discrete signal offset : ， When the baseband sampling clock frequency is Then the subsampling period delay for: ； Step 3.2 specifically includes the following steps: Substituting the sampling point positions and corresponding sampling point amplitudes of the sampling point data within the subsampling period delay calculation interval into the following probability density function model, and fitting the model, we obtain the corresponding fitted curve: ， Among them, among them, is the amplitude coefficient of the distribution, and is the parameter to be fitted; This represents the peak position, and the value is pending confirmation. is the scale parameter, corresponding to the half-width at half the maximum value of the distribution; The location of the sampling point; The signal amplitude distribution function; For signal amplitude distribution function By taking the derivative, we obtain its first derivative expression: ， The solution obtained The value represents the peak position. .

2. The high-precision timestamp detection method for wireless time synchronization according to claim 1, characterized in that, Obtain the reference time Includes the following steps: During the inflow of baseband discrete signals, the receiving end timestamp recording module (6) uses a state machine to monitor the entire frame data sequence of the frame header in real time. When the complete frame data sequence of the frame header is detected, the receiving end timestamp recording module (6) pulls up the Flag_demod signal and maintains it for two symbol periods. During the period when the Flag_demod signal is high, the receiving end timestamp recording module (6) detects the change in the received signal symbol bit and uses the timestamp of the first signal edge of the received signal symbol bit during the period when the Flag_demod signal is high as the reference time for data arrival. Reference time And the sampling point data sequence of the baseband discrete signal corresponding to the high level of the Flag_demod signal is input to the receiving end timestamp correction module (7), wherein the sampling point data sequence of the baseband discrete signal corresponding to the high level of the Flag_demod signal is the baseband discrete signal corresponding to the reference time. The corresponding sampling point data sequence.

3. The high-precision timestamp detection method for wireless time synchronization according to claim 1, characterized in that, The corrected arrival time is obtained based on the following formula: ， This indicates the corrected arrival time.

4. A high-precision timestamp detection device for wireless time synchronization, used to implement the high-precision timestamp detection method for wireless time synchronization as described in claim 1, characterized in that, Includes a transmitter and a receiver. The transmitter baseband processing module (1) converts the message into the transmitter baseband signal and inputs it into the transmitter timestamp module (2) and the transmitter radio frequency module (3). The transmitter timestamp module (2) detects and records the arrival time of the transmitter baseband signal as the transmission time t1. The transmitter radio frequency module (3) converts the transmitter baseband signal into a radio frequency signal for transmission. The radio frequency signal is received by the receiver's receiver radio frequency module (4) and converted into a baseband discrete signal before being sent to the frame decision module (5). The frame decision module (5) obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal through correlation calculations, and inputs the baseband discrete signal, the synchronization flag signal sync_flag, and the received signal symbol bit corresponding to the baseband discrete signal to the receiver timestamp recording module (6). The receiver timestamp recording module (6) performs frame detection on the baseband discrete signal in conjunction with the synchronization flag signal sync_flag, thereby obtaining the reference time of arrival of the baseband discrete signal corresponding to the message. and reference time and the baseband discrete signal at the reference time The corresponding sampling point data is input to the receiving end timestamp correction module (7); the receiving end timestamp correction module (7) uses the baseband discrete signal and the reference time The subsampling period delay is calculated based on the corresponding sampling point data, and the reference time is used. Make corrections to obtain the corrected arrival time. Meanwhile, the baseband discrete signal is input to the receiving baseband processing module (8) via the frame decision module (5) for demodulation.

Images