High-precision timestamp detection device and method for wireless time synchronization

Abstract

The invention discloses a high-precision timestamp detection device for wireless time synchronization. The high-precision timestamp detection device comprises a sending baseband processing module, a sending end timestamp module and a sending end radio frequency module, and a receiving end comprises a receiving end radio frequency module, a frame judgment module, a receiving end timestamp recording module, a receiving end timestamp correction module and a receiving baseband processing module. The invention also discloses a high-precision timestamp detection method for wireless time synchronization, and the method comprises the steps: processing a received signal through a correlation operation method, and obtaining a reference moment corresponding to the integral multiple sampling period time delay of the signal arrival time; and then fitting is carried out on the acquired baseband discrete signals by adopting a Lorentz distribution function to obtain sub-sampling period time delay, and the reference moment is further corrected, so that the precision of receiving the timestamp is effectively improved. The method effectively breaks through the problem that the traditional timestamp detection is limited by the sampling rate, improves the resolution of the received timestamp, and remarkably enhances the precision of the wireless time synchronization system.

Description

Technical Field

[0001] The present invention relates to the technical field of time synchronization, and in particular 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 Art

[0002] In wireless time synchronization, the timestamp serves as the core synchronization information for calculating clock deviations between nodes, and its accuracy has a decisive impact on the synchronization accuracy of the entire system. In comparison, obtaining the transmission timestamp is relatively simple, because the transmission time of the synchronization information is constrained by the channel access mechanism, and the physical layer processing delay is usually relatively fixed. Therefore, it is only necessary to record the local clock when the physical layer begins processing the synchronization information to obtain a more accurate transmission timestamp. In contrast, the accuracy of the reception timestamp depends mainly on the synchronization accuracy of the frame synchronization. Due to factors such as low signal-to-noise ratio environments and frequency offsets, the estimation performance of the frame arrival time is easily interfered with, which affects the accuracy of the reception timestamp. In a time synchronization system based on wireless communication, the process of obtaining the timestamp at the receiving end mainly relies on detecting the maximum value of the correlation peak of the physical layer data frame synchronization sequence.

[0003] Existing correlation time synchronization methods can be categorized as autocorrelation and cross-correlation. Autocorrelation methods utilize the inherent characteristics of the received signal to determine the time synchronization location, while cross-correlation methods locally generate a reference sequence identical to the transmitted sequence, perform a correlation operation 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 the signal is sampled and are therefore limited by the sampling clock. This means that the resolution of the received timestamp cannot be better than one sampling clock period, thus limiting the accuracy of time synchronization.

[0004] Therefore, the present invention proposes a high-precision timestamp detection method for wireless time synchronization. Based on a high-precision clock source, after completing the correlation operation, the sampled discrete signal is further subjected to Lorentz fitting, and then the fitted signal is subjected to differential operation, thereby realizing 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] The present invention belongs to the technical field of time and frequency, and provides a high-precision time stamp detection device for wireless time synchronization, and also provides a high-precision time stamp detection method for wireless time synchronization.

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

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

[0008] The RF signal is received by the RF module of the receiving end and converted into the baseband discrete signal of the receiving end and sent to the frame judgment module. The frame judgment module obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal through correlation operation, and inputs the baseband discrete signal, the synchronization flag signal sync_flag and the received signal symbol bit corresponding to the baseband discrete signal into the receiving end timestamp recording module; the receiving end timestamp recording module performs frame detection on the baseband discrete signal in combination with the synchronization flag signal sync_flag, thereby obtaining the reference time T of the arrival of the baseband discrete signal corresponding to the message rx_record , and the reference time T rx_record And the baseband discrete signal with reference time T rx_record The corresponding sampling point data is input to the receiving end time stamp correction module; the receiving end time stamp correction module uses the baseband discrete signal and the reference time T rx_record The corresponding sampling point data calculates the sub-sampling period delay and the reference time T rx_record Make corrections and get the corrected arrival time T rx_corrected At the same time, the baseband discrete signal is input into the receiving baseband processing module through the frame decision module for demodulation.

[0009] A high-precision timestamp detection method for wireless time synchronization, using 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 at the transmitting end sends the baseband signal of the transmitting end corresponding to the message to the timestamp module and the radio frequency module at the transmitting end; the timestamp module at 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 at the transmitting end converts the baseband signal of the transmitting end into a radio frequency signal and transmits it;

[0011] Step 2: The RF module at the receiving end receives the RF signal, converts the RF signal into a baseband discrete signal at the receiving end, and obtains the received signal sign bit corresponding to the baseband discrete signal. The decision module obtains the synchronization flag signal sync_flag corresponding to the baseband discrete signal through correlation operation;

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

[0013] The timestamp recording module at the receiving end 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 T of the arrival of the baseband discrete signal corresponding to the message. rx_record , and the baseband discrete signal with the reference time T rx_record The corresponding sampling point data sequence, the baseband discrete signal and the reference time T rx_record The corresponding sampling point data sequence is recorded as the sub-sampling period delay calculation interval;

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

[0015] Step 4: The timestamp correction module at the receiving end corrects the reference time T according to the sub-sampling period delay Δt. rx_record Make corrections and get the corrected arrival time T rx_corrected .

[0016] Get the reference time T as described above rx_record The steps include:

[0017] During the influx 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] When the Flag_demod signal is at a high level, the timestamp recording module at the receiving end detects the change of the received signal sign bit and uses the timestamp of the first signal edge of the received signal sign bit during the period when the Flag_demod signal is at a high level as the reference time of data arrival T rx_record , the reference time T rx_record The sampling point data sequence in the baseband discrete signal corresponding to the period when the Flag_demod signal is at a high level is input to the timestamp correction module at the receiving end, wherein the sampling point data sequence in the baseband discrete signal corresponding to the period when the Flag_demod signal is at a high level is the sampling point data sequence in the baseband discrete signal corresponding to the reference time T rx_record The corresponding sampling point data sequence.

[0019] Step 3 as described above includes the following steps:

[0020] Step 3.1: The timestamp correction module at the receiving end searches the sampling point data in the sub-sampling period delay calculation interval to find the maximum sampling point amplitude, and records the sampling point position corresponding to the maximum sampling point amplitude as the maximum discrete signal position x. pk ;

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

[0022] Step 3.3: Calculate the peak position x0 relative to the maximum discrete signal position x using the following formula: pk The offset δ:

[0023] δ=x0-x pk ,

[0024] When the baseband sampling clock frequency is f s , then the sub-sampling 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 Indicates the corrected arrival time.

[0029] The above step 3.2 specifically includes the following steps:

[0030] Substitute the sampling point positions and corresponding sampling point amplitudes of the sampling point data in the sub-sampling period delay calculation interval into the following probability density function model for fitting to obtain the corresponding fitting 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; f(x) is the signal amplitude distribution function;

[0033] Derivative the signal amplitude distribution function f(x) to obtain its first-order derivative expression:

[0034]

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

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

[0037] The present invention proposes a high-precision timestamp detection method for wireless time synchronization, which aims to improve the accuracy of the timestamp at the receiving end to improve the overall time synchronization performance. The method first uses the autocorrelation or cross-correlation principle based on a high-precision frequency source, combined with the output results of the correlation operation, to achieve a rough estimate of the integer multiple sampling period delay of the signal arrival time. On this basis, the Lorentz fitting model is introduced to fit the baseband discrete signal after the correlation operation, and then the fitted signal is differentiated to obtain the sub-sampling period delay. By performing sub-sampling level correction on the received timestamp, the influence of the sampling rate limitation on the synchronization accuracy is effectively overcome, providing key support for the construction of a high-precision, low-cost wireless time synchronization system.

[0038] The present invention adopts the Lorentz fitting waveform and performs differential solution on the fitting signal to make the obtained peak value more accurate, the fitting effect better and the precision higher:

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

[0040] Use multiple points to globally smooth the signal curve and effectively filter out local random noise;

[0041] By relying on function fitting, peak values ​​can be estimated even with "non-uniformly spaced" or "low sampling rate" data, avoiding fitting failures or offsets caused by fixed point spacing. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0043] Among them, 1-transmitting baseband processing module, 2-transmitting end timestamp module, 3-transmitting end RF module, 4-receiving end RF module, 5-frame judgment module, 6-receiving end timestamp recording module, 7-receiving end timestamp correction module, 8-receiving baseband processing module;

[0044] Figure 2 The timing diagram of the receiving end timestamp recording module algorithm; S0-S32 represents the 32 bits of the frame header, corresponding to 33 states of the state machine; filtered_sig refers to the received signal sign 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 reference time T rx_record ;

[0045] Figure 3 A partial enlarged diagram of the algorithm for recording timestamps at the receiving end;

[0046] Figure 4 This is the Lorentz distribution fitting diagram of the timestamp correction module at the receiving end;

[0047] Figure 5 The derivative function diagram of the Lorentz distribution fitting of the timestamp correction module at the receiving end. DETAILED DESCRIPTION

[0048] In order to facilitate those skilled in the art to understand and implement the present invention, the present invention is further described in detail below in conjunction with embodiments. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention and are not used 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 transmitting end and a receiving end. The transmitting end includes a transmitting baseband processing module 1, a transmitting end timestamp module 2, and a transmitting end radio frequency module 3. The receiving end includes a receiving end radio frequency module 4, a frame judgment module 5, a receiving end timestamp recording module 6, a receiving end timestamp correction module 7 and a receiving baseband processing module 8. The above modules are all implemented by FPGA.

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

[0052] The receiving end RF module 4 receives the RF signal, converts the RF signal into a baseband discrete signal at the receiving end, and sends it to the frame decision module 5. The baseband discrete signal is a discrete sampling point data sequence; 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 operation, 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 receiving end timestamp recording module 6; the receiving end timestamp recording module 6 performs frame detection on the baseband discrete signal in combination with the synchronization flag signal sync_flag, thereby detecting and recording the reference time T of the arrival of the baseband discrete signal corresponding to the message rx_record , and the reference time Trx_record And the baseband discrete signal with reference time T rx_record The corresponding sampling point data is input to the receiving end time stamp correction module 7; the receiving end time stamp correction module 7 uses the baseband discrete signal and the reference time T rx_record The corresponding sampling point data estimates the sub-sampling period delay, and the reference time T is calculated based on the sub-sampling period delay. rx_record Make corrections and get the corrected arrival time T rx_corrected At the same time, 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 received baseband processing module 8 receives the corrected arrival time T rx_corrected Embedded into 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, using a 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. Among them, the estimation of the integer multiple sampling period delay is used to preliminarily determine the reference time T of arrival of the baseband discrete signal at the receiving end corresponding to the signal message. rx_record The estimation of sub-sampling period delay is used to further accurately locate the real time of signal arrival and realize timestamp correction with sub-sampling accuracy.

[0056] It should be noted that the present invention is applicable to both autocorrelation and cross-correlation synchronization methods. While the two methods differ slightly in how integer multiple delays are obtained, they are consistent in how fractional sampling period delays are estimated. For ease of explanation, this embodiment uses the autocorrelation method employed by the receiving end as an example for detailed description.

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

[0058] Step 1: Detect the sending timestamp:

[0059] The transmitting baseband processing module 1 of the transmitting end sends the baseband signal of the transmitting end corresponding to the message to the transmitting end timestamp module 2 and the transmitting end radio frequency module 3; the transmitting end timestamp module 2 detects and records the arrival time of the baseband signal of the transmitting end as the transmitting time t1; the transmitting end 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 check, timestamp data, first timestamp data check and second timestamp data check, wherein the type is used to indicate which frame message it is, wherein 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 reached by the receiving end signal;

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

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

[0062] In the receiving end timestamp recording module 6, the core problem is how to accurately record the arrival time of the valid data at the receiving end (that is, the baseband discrete signal corresponding to the message). In order to ensure the validity of the timestamp of the baseband discrete signal arrival time corresponding to the message, the receiving end timestamp recording module 6 combines the synchronization flag signal sync_flag and the frame header data pattern of the message to perform frame detection on the baseband discrete signal to obtain the reference time Trx_record of the valid data arrival at the receiving end. 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 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 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). During the period when the Flag_demod signal is at a high level, the receiving end timestamp recording module 6 detects changes in the sign bit of the received signal (in this embodiment, changes in the sign bit of the received signal of the in-phase branch signal filtered_I or the orthogonal branch signal filtered_Q are detected), and timestamps the moment of the first signal edge of the sign bit of the received signal during the period when the Flag_demod signal is at a high level (this embodiment selects the rising edge or the falling edge 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 is recorded as the reference time T of data arrival. rx_record The solution of the present invention uses the signal edge of the received signal sign bit as the reference signal to obtain the reference time T rx_record The process no longer relies on traditional synchronization marker signals, effectively avoiding the errors and uncertainties caused by synchronization delays, thereby further improving the accuracy of time synchronization.

[0064] The receiving end timestamp recording module 6 uses the reference time T rx_record The sampling point data sequence in the baseband discrete signal corresponding to the period when the Flag_demod signal is at a high level is input to the receiving end timestamp correction module 7, wherein the sampling point data sequence in the baseband discrete signal corresponding to the period when the Flag_demod signal is at a high level is used as the sampling point data sequence in the baseband discrete signal corresponding to the reference time T rx_record The corresponding sampling point data sequence is recorded as the sub-sampling 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 sub-sampling period delay of the receiving end timestamp:

[0067] In the timestamp detection at the receiving end, in order to break through the sampling rate limit and improve the accuracy of the timestamp, it is necessary to estimate the sub-sampling period delay between sampling points. The sub-sampling period delay reflects the difference between the actual arrival time of the signal and the reference time T rx_recordThe deviation between the two is the key to achieving sub-sampling precision time synchronization. Specifically, when the Flag_demod signal, which serves as the synchronization flag, is detected to be high, indicating that the baseband discrete signal corresponding to the message has arrived, the receiving end timestamp correction module 7 immediately calculates the sub-sampling period delay based on the sampling point data in the sub-sampling 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 in the sub-sampling period delay calculation interval (such as Figure 3 As shown in Figure 2, the sampling point data includes the sampling point position and the sampling point amplitude, and the sampling point position corresponding to the maximum sampling point amplitude is recorded as the maximum discrete signal position x. pk , the sampling point position is the sampling point sequence number.

[0069] Step 3.2: To more accurately estimate the true position of the maximum amplitude value in the sub-sampling period delay calculation interval and obtain a time estimate that exceeds the sampling period accuracy, the sampling point data in the sub-sampling period delay calculation interval is fitted based on the Lorentzian distribution function to obtain the peak position x0 of the sub-sampling period delay calculation interval.

[0070] Substitute the sampling point position and the corresponding sampling point amplitude of the sampling point data in the sub-sampling period delay calculation interval into the following probability density function model for fitting to obtain the corresponding fitting curve. The fitting result is as follows: 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, corresponding 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 sub-sampling period delay calculation interval is substituted into x, and the sampling point amplitude is substituted into f(x; x0; γ).

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

[0074]

[0075] The above derivative function reflects the change of the slope of the fitting curve, has the characteristics of an odd function, and takes a value of zero at x = x0, that is, the x value obtained by solving f'(x) = 0 is the peak position x0. The derivative image is as follows Figure 5As shown in Figure 3, the zero point position clearly reflects the signal peak location, thus enabling high-precision estimation of the arrival time.

[0076] Step 3.3: Obtain the peak position x0 relative to the maximum discrete signal position x limited by the integer sampling period using the following formula: pk The offset δ:

[0077] δ=x0-x pk ,

[0078] When the baseband sampling clock frequency is f s , we can get the sub-sampling period delay Δt:

[0079]

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

[0081] If δ>0, it means that the locally recorded data reaches the reference time T rx_record It is earlier than the real signal arrival time, so it should be corrected backwards:

[0082] T rx_corrected =T rx_record +Δt,

[0083] T rx_corrected Indicates the corrected arrival time.

[0084] If δ<0, it means that the local recorded timestamp lags behind the real signal arrival time and needs to be corrected forward:

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

[0086] Through the above method, the receiving timestamp can achieve compensation and correction of the fractional sampling period, thereby repairing the time synchronization error introduced by the limited sampling clock accuracy, and further achieving the time synchronization error within one sampling period.

[0087] By using a curve fitting method, this invention effectively overcomes the technical bottleneck of traditional correlation detection, which is limited by the sampling rate and difficulty in achieving sub-sampling period accuracy. Furthermore, this method is applicable to both autocorrelation and cross-correlation synchronization methods. While these two methods differ in how they obtain integer multiples of the sampling period delay, they maintain consistency in their sub-sampling period delay estimation methods.

[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 based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A high-precision timestamp detection device for wireless time synchronization, characterized in that: The invention comprises a transmitting end and a receiving end, wherein a transmitting baseband processing module (1) of the transmitting end converts a message into a baseband signal of the transmitting end and inputs the message into a transmitting end time stamp module (2) and a transmitting end radio frequency module (3), wherein the transmitting end time stamp module (2) detects and records the arrival time of the transmitting end baseband signal as the transmitting time t1, and the transmitting end radio frequency module (3) converts the transmitting end baseband signal into a radio frequency signal for transmission; The radio frequency signal is received by a receiving end radio frequency module (4) at the receiving end, and is converted into a baseband discrete signal at the receiving end and then sent to a frame decision module (5). The frame decision module (5) obtains a synchronization flag signal sync_flag corresponding to the baseband discrete signal through a correlation operation, and inputs the baseband discrete signal, the synchronization flag signal sync_flag, and the received signal symbol bit corresponding to the baseband discrete signal into a receiving end timestamp recording module (6). The receiving end timestamp recording module (6) performs frame detection on the baseband discrete signal in combination with the synchronization flag signal sync_flag, thereby obtaining a reference time T at which the baseband discrete signal corresponding to the message arrives. rx_record , and the reference time T rx_record And the baseband discrete signal with reference time T rx_record The corresponding sampling point data is input to the receiving end time stamp correction module (7); the receiving end time stamp correction module (7) uses the baseband discrete signal and the reference time T rx_record The corresponding sampling point data calculates the sub-sampling period delay and the reference time T rx_record Make corrections and get the corrected arrival time T rx_corrected At the same time, the baseband discrete signal is input to the receiving baseband processing module (8) through the frame decision module (5) for demodulation.

2. A high-precision timestamp detection method for wireless time synchronization, using the high-precision timestamp detection device for wireless time synchronization according to claim 1, characterized in that: The following steps are involved: Step 1: The transmitting baseband processing module (1) of the transmitting end sends the baseband signal of the transmitting end corresponding to the message to the transmitting end timestamp module (2) and the transmitting end radio frequency module (3); the transmitting end timestamp module (2) detects and records the arrival time of the baseband signal of the transmitting end as the transmitting time t1; the transmitting end radio frequency module (3) converts the baseband signal of the transmitting end into a radio frequency signal and transmits it; Step 2: The receiving end radio frequency module (4) receives the radio frequency signal, converts the radio frequency signal into a baseband discrete signal of the receiving end and obtains a received signal sign bit corresponding to the baseband discrete signal. The decision module (5) obtains a synchronization flag signal sync_flag corresponding to the baseband discrete signal through a correlation operation; The frame decision module (5) inputs the baseband discrete signal, the synchronization flag signal sync_flag and the received signal symbol bit into the receiving end timestamp recording module (6); The receiving end 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 T of arrival of the baseband discrete signal corresponding to the message. rx_record , and the baseband discrete signal with the reference time T rx_record The corresponding sampling point data sequence, the baseband discrete signal and the reference time T rx_record The corresponding sampling point data sequence is recorded as the sub-sampling period delay calculation interval; Step 3, the receiving end timestamp correction module (7) calculates the sub-sampling period delay Δt based on the sampling point data in the sub-sampling period delay calculation interval; Step 4: The receiving end timestamp correction module (7) adjusts the reference time T according to the sub-sampling period delay Δt. rx_record Make corrections and get the corrected arrival time T rx_corrected .

3. A high-precision timestamp detection method for wireless time synchronization according to claim 2, characterized in that: Get the reference time T rx_record The steps include: During the inflow of the baseband discrete signal, 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 at a high level, the receiving end timestamp recording module (6) detects the change of the received signal sign bit, and uses the timestamp of the first signal edge of the received signal sign bit during the period when the Flag_demod signal is at a high level as the reference time T of data arrival. rx_record , the reference time T rx_record The sampling point data sequence in the baseband discrete signal corresponding to the period when the Flag_demod signal is at a high level is input to the receiving end timestamp correction module (7), wherein the sampling point data sequence in the baseband discrete signal corresponding to the period when the Flag_demod signal is at a high level is the sampling point data sequence in the baseband discrete signal corresponding to the reference time T rx_record The corresponding sampling point data sequence.

4. The high-precision timestamp detection method for wireless time synchronization according to claim 2, characterized in that: The step 3 comprises the following steps: Step 3.1, the receiving end timestamp correction module (7) searches the sampling point data in the sub-sampling period delay calculation interval to find the maximum sampling point amplitude, and records the sampling point position corresponding to the maximum sampling point amplitude as the maximum discrete signal position x pk ; Step 3.2: Fit the sampling point data in the sub-sampling period delay calculation interval based on the Lorentz distribution function to obtain the peak position x0 of the sub-sampling period delay calculation interval; Step 3.3: Calculate the peak position x0 relative to the maximum discrete signal position x using the following formula: pk The offset δ: δ=x0-x pk , When the baseband sampling clock frequency is f s , then the sub-sampling period delay Δt is:

5. A high-precision timestamp detection method for wireless time synchronization according to claim 4, characterized in that: The corrected arrival time is obtained based on the following formula: T rx_corrected =T rx_record +Δt, T rx_corrected Indicates the corrected arrival time.

6. A high-precision timestamp detection method for wireless time synchronization according to claim 4, characterized in that: The step 3.2 specifically includes the following steps: Substitute the sampling point positions and corresponding sampling point amplitudes of the sampling point data in the sub-sampling period delay calculation interval into the following probability density function model for fitting to obtain the corresponding fitting curve: 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; Derivative the signal amplitude distribution function f(x) to obtain its first-order derivative expression: The x value obtained by solving f′(x)=0 is the peak position x0.

Images