Time-frequency synchronization method and system, electronic equipment and storage medium

Abstract

The invention discloses a time-frequency synchronization method and system, electronic equipment and a storage medium, and relates to the technical field of communication. The method comprises the following steps: acquiring a received signal, and performing iterative equivalent matched filtering on the received signal based on symbol delay to obtain a burst signal preliminary positioning result; and generating a local training symbol based on a preset frequency domain spread spectrum sequence, and performing parallel matched filtering by using the local training symbol and the initial positioning result of the burst signal to obtain a joint detection result of frequency offset estimation and time domain synchronization. Through innovative iteration equivalent matching detection, frequency domain sequence design and parallel matching processing, the carrier synchronization performance of the OFDM satellite communication system is greatly improved, synchronization is faster, frequency offset estimation is more accurate, coverage frequency offset is larger, weak signal adaptability is higher, meanwhile, implementation complexity and resource overhead are reduced, and the method is suitable for large-scale popularization and application. And the method has obvious engineering application value.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a time-frequency synchronization method, system, electronic device, and storage medium for a low-Earth orbit satellite burst broadband OFDM system. Background Technology

[0002] Currently, in OFDM communication scenarios, receivers typically utilize sliding window energy to achieve OFDM frame detection when there is no prior information. While this method improves frame detection performance to some extent, it suffers from limitations such as the power of the received signal easily being confused with the noise power when the signal-to-noise ratio is low. Therefore, it has poor robustness in interference-prone communication scenarios, and the accuracy and speed of frame detection need to be improved. For bursty OFDM signals with large duty cycles, due to their short time-domain duration, broadened spectrum, and complex energy distribution, common methods such as point-by-point sliding correlation matching and time-frequency joint detection require intensive computation within the entire sampling window. This not only requires a large amount of correlation calculation and accumulation operations but also leads to excessive processor resource consumption, significant detection delay, and difficulty in ensuring real-time performance, resulting in low overall detection efficiency.

[0003] Furthermore, existing OFDM communication carrier synchronization methods typically rely on the maximum likelihood estimation (ML) algorithm for cyclic prefixes to estimate the fractional frequency offset (FFO), and then use subcarrier offsets in the frequency domain to estimate the integer frequency offset (IFO). Alternatively, they may use specially constructed pilot symbols to achieve time-frequency synchronization. While these existing methods can achieve synchronization, they still suffer from drawbacks such as low efficiency and accuracy in burst signal detection, limited time-frequency estimation precision, and limited noise immunity. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the above-mentioned defects in the prior art and provide a time-frequency synchronization method, system, electronic device, and storage medium based on burst signal detection.

[0005] The present invention solves the above-mentioned technical problems through the following technical solution: Firstly, a time-frequency synchronization method is provided, including: The received signal is acquired, and iterative equivalent matched filtering is performed on the received signal based on the symbol delay to obtain the preliminary location result of the burst signal; Local training symbols are generated based on a preset frequency domain spread spectrum sequence. The local training symbols and the preliminary location results of the burst signal are used for parallel matched filtering to obtain joint synchronization detection results.

[0006] Secondly, a time-frequency synchronization system is provided, comprising: The burst signal localization module is used to acquire the received signal and perform iterative equivalent matched filtering on the received signal based on the symbol delay to obtain the preliminary localization result of the burst signal; The joint synchronization module is used to generate local training symbols based on a preset frequency domain spread spectrum sequence, and to perform parallel matched filtering using the local training symbols and the preliminary location results of the burst signal to obtain the joint synchronization detection results.

[0007] Thirdly, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in any of the preceding claims.

[0008] Fourthly, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method described in any of the preceding claims.

[0009] The beneficial effects of this invention are as follows: This invention significantly improves the performance of carrier synchronization in OFDM satellite communication systems through innovative burst signal detection, iterative equivalent matching detection, frequency domain sequence design, and parallel matching processing: synchronization is faster, frequency offset estimation is more accurate, coverage frequency offset is larger, and weak signal adaptability is stronger. At the same time, the complexity and resource overhead are reduced, which has significant engineering application value. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating a time-frequency synchronization method provided as an exemplary embodiment of the present invention.

[0011] Figure 2 This is a schematic diagram of the result of iterative equivalent matched filtering provided for an exemplary embodiment of the present invention.

[0012] Figure 3 This is a schematic diagram of the result of continuous correlation matching provided for an exemplary embodiment of the present invention.

[0013] Figure 4 A schematic diagram of the correct integer multiple subcarrier frequency offset matching result provided for an exemplary embodiment of the present invention.

[0014] Figure 5 This is a schematic diagram of the frequency offset matching result of an erroneous integer multiple subcarrier provided as an exemplary embodiment of the present invention.

[0015] Figure 6 A schematic diagram illustrating the phase effect of fractional multiple subcarrier frequency offset as provided in an exemplary embodiment of the present invention.

[0016] Figure 7 This is a schematic diagram of the structure of a time-frequency synchronization system shown in an example embodiment of the present invention. Detailed Implementation

[0017] The present invention will be further illustrated by way of embodiments below, but the present invention is not limited to the scope of the embodiments described herein.

[0018] This invention proposes a time-frequency synchronization method based on the detection of burst OFDM signals from satellites. First, a pair of window functions (dual-window filters) with specific delays are designed using known signal frame structure information. The delay is specified as a multiple of the period of the repetitive signal in the signal frame, and an asynchronous long-interval iterative strategy is employed for equivalent matched filtering. This method efficiently detects burst OFDM signals with large duty cycles and quickly pinpoints their approximate starting position in the time domain, completing initial burst signal acquisition. After initial burst signal acquisition, the process enters the frequency domain processing stage: first, a pre-designed frequency domain spreading sequence is inserted at the transmitter; at the receiver, the frequency domain spreading sequence is used to perform multiple sets of hypothetical matched filtering (i.e., correlation matching of multiple possible combinations of synchronization parameters) to achieve accurate and rapid detection of integer frequency offset and time domain synchronization. Subsequently, the fractional frequency offset can be further accurately estimated by measuring the phase-frequency offset characteristics of the correlation peak.

[0019] Figure 1 This is a schematic flowchart of a time-frequency synchronization method provided in an exemplary embodiment of the present invention. The following embodiments are combined with... Figure 1 The specific methods will be explained.

[0020] Step s01: Perform iterative equivalent matched filtering based on symbol delay on the received signal to quickly detect burst signals in the received signal in the time domain and obtain preliminary location of burst signals.

[0021] After acquiring the received signal, the first step is to detect any burst OFDM signals it contains. This invention utilizes the similarity between different OFDM symbols and pilot repetition characteristics, employing a sliding dual-window matching calculation delayed by several complete OFDM symbol periods to achieve rapid detection of burst frames and enhance the signal-to-noise ratio, thereby quickly locating burst signals in the received signal in the time domain.

[0022] Compared to traditional sliding correlation methods, the improved method in this embodiment of the invention does not require full-length correlation for each sampling point, but only requires a single delay inner product to obtain the equivalent matching result, thereby significantly reducing computational complexity and improving detection real-time performance. Since this detection method utilizes the matching relationship of the signal itself, the improved method in this embodiment is called Equivalent Matched Filtering (EMF).

[0023] Specifically, let the length of a single OFDM symbol (including the cyclic prefix) be... Two time windows are constructed by sliding the received signal (constant sampling points) along the time axis with fixed or variable steps, and the delay between the two time windows is such that... It is equal to an integer multiple of the period of a complete OFDM symbol, that is k=1,2,3,...... and multiples The value of is less than the total number of OFDM symbols, num.

[0024] The signals within the corresponding two time windows are denoted as follows: (1) (2) in It is a complex baseband discrete sampling sequence. For window length, This is the index for discrete sampling points.

[0025] The time-domain model of the sampled signal (with CP removed) can be expressed as: (3).

[0026] Where δ is the normalized subcarrier spacing. Indicates noise. This indicates the actual delay.

[0027] In a preferred embodiment, the window length of the two time windows can be one or more OFDM symbol periods to increase the signal-to-noise ratio. The specific value can be set according to the scenario conditions. This embodiment of the invention does not impose a specific limitation on the window length.

[0028] Define equivalent matched filter output for: (4) Simultaneously define the reference power result for (5) Calculated continuously in the time domain with compensation and This allows us to obtain the burst signal detection curve. Specifically, when the delay... When the OFDM symbol length is an integer multiple of the total OFDM symbol length, if the repeated pilots or structural portions between adjacent OFDM symbols overlap within the double window, then the following conditions must be met: (6) in The overall phase rotation that may exist between OFDM symbols is caused by factors such as carrier frequency offset. Substituting this into formula (4) yields: (7) in The window signal energy indicates that the matching result is only related to the delay. Related to, but not related to, specific discrete sampling point indexes The position is irrelevant. Therefore, as long as the two windows fall completely within the symbol range of the burst OFDM signal, a stable and significantly significant matched peak value can be obtained. However, when the window is located in a no-signal or noisy region, since the two signal segments corresponding to the two windows are uncorrelated, the noise cannot be effectively accumulated, and the output result is only random noise fluctuation.

[0029] Further considering the presence of additive white Gaussian noise in the received signal, since white Gaussian noise is uncorrelated under different delays, the noise energy cannot accumulate. Meanwhile, the repetitive signal components between symbols undergo coherent superposition under the matched delay, resulting in an output signal-to-noise ratio increase of approximately times the window length. This indicates that equivalent matched filtering (EMF) detection can significantly improve the detection signal-to-noise ratio while maintaining low complexity, making the burst frame initiation still detectable under low signal-to-noise ratio conditions.

[0030] Furthermore, in a preferred embodiment, a two-step variable-step iterative skip search strategy can be employed to obtain preliminary localization results for the burst signal: firstly, a coarse search is performed in the time domain with the first step size to quickly locate the time segment where a matching peak may exist; after the peak is detected, a fine search is performed in its neighborhood with the second step size to accurately determine the starting position of the burst OFDM signal. Since each slide involves only one delay inner product calculation, the computational load is linearly related to the number of searches, and the overall complexity is much lower than that of the traditional full correlation detection algorithm, ultimately forming Iterative Equivalent Matched Filtering (IEMF) detection.

[0031] Furthermore, the first step length is no more than the length of the burst signal pilot repetition structure. The second step length is the length of a single OFDM symbol.

[0032] Furthermore, the step size of the second step can be continuously shortened and refined within the range of a single OFDM symbol length with each iteration of detection. However, the increase in the number of detection calculations brought about by refining the step size needs to be considered. Therefore, in practical applications, the specific step size of the second step can be selected by combining the two factors.

[0033] Figure 2 The detection performance of Iterative Equivalent Matched Filtering (IEMF) on real sampled signals is shown in the figure. Figure 2 The left image shows an example of a real sampled signal. Figure 2 The middle figure shows the signal power intensity corresponding to the actual sampled signal. Figure 2The right figure shows the preliminary localization result of the burst signal obtained by iterative equivalent matched filtering (IEMF) detection of the real sampled signal. As can be seen, the improved iterative equivalent matched filtering (IEMF) detection method in this embodiment can obtain a gain much greater than the signal strength, has a better detection effect, and does not require traditional point-by-point sliding correlation, thus maintaining good detection performance at low signal-to-noise ratios.

[0034] Step s02: Based on the preset frequency domain spread spectrum sequence and different assumed subcarrier offsets, generate local training symbols. Use the local training symbols to perform parallel matched filtering within the range of the initial location results of the burst signal to obtain the joint synchronization detection result of frequency offset estimation and time domain synchronization.

[0035] For unknown integer multiples of carrier frequency offset, this embodiment further employs a parallel local reference matching method with multiple assumed subcarrier frequency offsets: after obtaining the preliminary location result of burst signals in the received signal through iterative equivalent matched filtering (IEMF), the frequency domain spreading sequence corresponding to the frequency domain spreading code is spread in the frequency domain according to different assumed subcarrier offsets. Perform cyclic shifting (corresponding to compensation for different numbers of assumed subcarrier biases) Specifically, it is assumed that the number of subcarrier offsets can be obtained by dividing the maximum downlink Doppler frequency of the satellite signal by the subcarrier frequency interval. After passing through a Fast Fourier Transform (IFFT), multiple sets of offset reference sequences can be obtained simultaneously. In other words, it is assumed that frequency offset may cause the received sequence to be cyclically shifted by several subcarrier offsets. Then the bias reference sequence will also be shifted accordingly, assuming a subcarrier bias. The signal is aligned to this assumption. Subsequently, each bias reference sequence is converted into a local training symbol, and a sliding correlation operation is performed with the preliminary localization result of the burst signal to achieve parallel matched filtering.

[0036] For example, the generation of local training symbols, Figure 3 As shown, the receiver pre-stores or generates reference symbols identical to the frequency domain spreading code c used by the transmitter, based on preset information, as the frequency domain spreading sequence c for sliding correlation matching, expressed as: .

[0037] Where N represents the number of FFT / IFFT points for a single symbol.

[0038] In a preferred embodiment, the different assumed subcarrier offsets are set by combining the maximum Doppler frequency and the subcarrier spacing in the usage scenario. For example, if the subcarrier spacing is 100kHz and the maximum Doppler frequency is 200kHz, then the assumed subcarrier offsets are -2, -1, 0, 1, and 2, totaling 5 possibilities. The subcarrier index set can be represented as... .

[0039] When using local integer multiples, assume subcarrier offset Then the local training symbol x corresponding to the frequency domain spread spectrum sequence c in the time domain local Represented as: (8).

[0040] Where s represents the amplitude and k represents the normalized frequency offset.

[0041] In a preferred embodiment, the correlation operation can be performed within a window of one OFDM symbol length: the local training symbols corresponding to all assumed subcarrier offsets are multiplied point-by-point by conjugate with the window containing the initial burst signal localization results, and then accumulated to obtain a sequence of correlation values ​​under the corresponding offset assumptions. Through parallel operations or fast cyclic correlation, matching of all possible integer offset values ​​within a preset frequency offset search range is achieved. In one embodiment, the preset frequency offset search range is the maximum frequency offset range determined after burst signal acquisition.

[0042] This process is equivalent to searching in two dimensions: time and frequency offset. In the time dimension, the position of the sliding window determines the initial alignment of the OFDM symbol, while in the frequency dimension, switching the local offset determines the assumed integer frequency offset IFO correction. Ultimately, the correlation output will form a two-dimensional power peak, reaching its maximum under correct time alignment and correct frequency offset assumptions.

[0043] Among them, "correct time alignment" means that the local training symbols in the time domain are completely consistent with the starting sampling point of the received signal in the time domain; "correct frequency offset assumption" means assuming subcarrier offset. The subcarrier offset m is consistent with the actual integer multiple of the received signal. When both conditions are met, the two-dimensional matched output reaches its global maximum peak.

[0044] like Figure 3 As shown, under frequency dimension matching, the result of sliding matching of local training symbols with continuous OFDM signals in the time domain will be a series of equally spaced correlation peaks. Each group of correlation peaks consists of a main peak and two sub-peaks on the left and right. The main peak is obtained by matching the entire OFDM signal, and the sub-peaks are obtained by matching the CP signal. Therefore, the interval between the main peaks is the length of the OFDM symbol, and the interval between the sub-peaks and the main peaks is the length of the CP signal.

[0045] By performing a sliding cross-correlation on each integer-delayed sampling point, the correlation function expression can be obtained as follows: (9) in For subcarrier index, For the time-domain discrete sampling point delay, Indicates conjugate.

[0046] The corresponding detection function (i.e., normalized matching energy) can be expressed as: (10) in This represents the total energy of the locally trained symbols.

[0047] In the actual synchronization process, the receiving end first detects the global maximum correlation peak in the aforementioned correlation results. The peak value detection result of the global maximum correlation peak. For a given assumed subcarrier bias and the starting point of a certain sliding window This means that when the local training symbol is frequency-shifted by the assumed subcarrier offset and aligned with that moment, the match with the received signal is the highest.

[0048] From this, we can obtain the following results: (1) Integer multiple carrier frequency offset (IFO) estimation: Based on the subcarrier frequency offset hypothesis that produces the maximum correlation peak The carrier frequency offset, which is directly determined to be an integer multiple of the received signal, is estimated as follows: , The subcarrier spacing frequency.

[0049] For example, if the correlation peak reaches its global maximum value at a local offset of +3 subcarriers, then the integer frequency offset is determined to be +3. In this way, by scanning the position of the correlation peak, the integer multiple carrier offset can be found in one step.

[0050] (2) Symbol time synchronization: The location of the global maximum correlation peak within a given time window is the starting position of the synchronous training symbols.

[0051] Since the correlation operation covers the entire possible symbol arrival interval, the location of the maximum correlation peak indicates that the OFDM symbol boundary has been successfully captured. Therefore, accurate frame synchronization / symbol timing is achieved simultaneously with obtaining the integer frequency offset estimate (IFO).

[0052] For burst signals containing multiple OFDM symbols, Figure 4 and Figure 5 The results show the correct integer multiple subcarrier frequency offset matching and the incorrect integer multiple subcarrier frequency offset matching, respectively. Figure 4 and Figure 5 The comparison reveals the time-domain matching results of the local training symbols: for correct subcarrier offsets, the matching results show high-intensity, equally spaced correlation peaks, while for incorrect subcarrier offsets, there are no obvious correlation peaks. Therefore, this can be used as a criterion to determine whether the subcarrier matching is correct.

[0053] (3) Fractional Frequency Offset (FFO) estimation: After obtaining the integer frequency offset estimate (IFO), this embodiment of the invention further utilizes the phase characteristics of the correlation output to extract the remaining fractional frequency offset. .

[0054] When integer multiples of frequency offset If the maximum correlation peak has already been determined, let the correlation peak value be:

[0055] in It is the complex phase angle of the peak value.

[0056] Specifically, even after frequency shift correction to integer multiples of subcarriers, a residual frequency difference of less than one subcarrier interval may still exist, making the correlation peak a complex number rather than a purely real number. The complex phase of the correlation peak reflects the relative phase offset that still exists between the local reference signal and the received signal, and it is proportional to the fractional frequency offset. By calculating the phase angle of this peak, the fractional frequency offset can be estimated.

[0057] When the assumed local frequency offset is lower than the actual frequency offset, the phase of the correlation peak is positive (leading); when it is higher than the actual frequency offset, the phase is negative (lagging). Based on the linear relationship between phase and offset, the offset amount that makes the phase zero can be interpolated, i.e., the corresponding fractional frequency offset value. At this point, the residual fractional frequency offset... It is directly proportional to the phase angle and can be expressed as: .

[0058] in This refers to the OFDM symbol duration.

[0059] Figure 6 We can see the effective correlation peak phase change when using the frequency domain spread spectrum training sequence and the sampled signal sliding match. This phase change is caused by the residual fractional multiples of the subcarrier frequency. By fitting the phase change, the rate of change (slope) can be used as an estimate of the residual fractional multiples of the subcarrier frequency offset.

[0060] In summary, by utilizing the phase information of relevant peak values, the solution provided by this embodiment of the invention can complete the estimation of fractional frequency offset (FFO) without additional pilot signals.

[0061] The final estimated Doppler frequency is an integer multiple of the subcarrier spacing frequency plus a fractional multiple of the subcarrier spacing frequency.

[0062] As can be seen, this embodiment of the invention performs parallel correlation calculations for various frequency offset assumptions, and automates the search for large-range frequency offsets by obtaining subcarrier frequency offsets through cyclic shifting in the frequency domain instead of repeatedly sliding the spectrum. This method effectively reduces the complexity of large-range frequency offset search and ensures real-time performance. This parallel matched filtering architecture is also a key innovation of this embodiment, enabling the acquisition of complete synchronization information in a single training iteration.

[0063] Corresponding to the aforementioned time-frequency synchronization method embodiments, the present invention also provides an embodiment of a time-frequency synchronization system.

[0064] Figure 7 This is a schematic diagram of a time-frequency synchronization system based on burst signal detection, provided as an exemplary embodiment of the present invention. The time-frequency synchronization system is used to execute the time-frequency synchronization method described in any of the foregoing embodiments.

[0065] Specifically, the time-frequency synchronization system includes, The burst signal localization module 71 is used to acquire the received signal, perform iterative equivalent matching based on symbol delay on the received signal, and obtain the preliminary localization result of the burst signal. The joint synchronization module 72 is used to generate local training symbols based on a preset frequency domain spread spectrum sequence, and to perform parallel matched filtering using the local training symbols and the preliminary location results of the burst signal to obtain joint synchronization detection results.

[0066] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative, and the modules described as separate modules may or may not be physically separate. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present invention. Those skilled in the art can understand and implement this without creative effort.

[0067] Another embodiment of the present invention also provides a corresponding electronic device. This electronic device is merely an example and should not be construed as limiting the functionality or scope of the embodiments of the present invention. The electronic device can be in the form of a general-purpose computing device, such as a server device. The components of the electronic device may include, but are not limited to: at least one processor, at least one memory, and buses connecting different system components. The buses include data buses, address buses, and control buses.

[0068] The memory may include program tools having at least one program module, including but not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. The processor executes various functional applications and data processing, such as the methods provided in any of the above embodiments, by running computer programs stored in the memory.

[0069] It should be noted that although several units / modules or sub-units / modules of the electronic device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of the present invention, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.

[0070] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method provided in any of the above embodiments.

[0071] It should be noted that although the process of this invention uses multi-reference parallel correlation, its computational structure is well-organized and its data path is clear, making it very suitable for pipelined implementation in field-programmable gate arrays or on-chip systems. The core components can all be assembled using mature digital signal processing units. More importantly, when the system needs to expand to a wider frequency offset search or support multiple transmit / receive diversity, it can be linearly expanded without changing the overall control logic simply by horizontally replicating the reference pipeline or increasing the parallelism. This smooth hardware expansion path leaves ample room for future iterations and facilitates the rapid deployment of the algorithm to terminal platforms with different performance levels, demonstrating good engineering adaptability.

[0072] In summary, the improved scheme of this invention has the advantages of higher efficiency and accuracy in burst signal detection. Compared with the prior art, this invention achieves cross-symbol equivalent matched filtering by introducing an inter-symbol delay matching mechanism, which significantly reduces computational complexity while achieving coherent accumulation of signal energy and non-accumulation of noise energy. Thus, it can still generate stable detection peaks under low signal-to-noise ratio conditions, and has higher detection sensitivity, real-time performance, and robustness.

[0073] Secondly, the improved scheme of this invention offers faster time-frequency synchronization and stronger robustness under weak signal conditions. Compared to traditional methods that require sequential execution of three major steps—coarse timing, rapid transformation, and frequency offset detection—this invention employs a parallel correlation search framework, mapping the time offset and integer frequency offset to the same power map in one go, completing all coarse synchronization steps in just one symbol period. Simultaneously, it utilizes a full-band energy of one symbol length during the correlation accumulation phase, working in conjunction with parallel pipelines to simultaneously calculate multiple hypotheses, resulting in higher signal-to-noise gain at the peak. This gain reduces the probability of the receiver falling into prolonged blind scans when satellite coverage is at the edge or when severe weather causes significant link attenuation. For airborne or vehicle-mounted high-speed mobile terminals, it directly shortens the blind zone time waiting for channel establishment. Even on handheld terminals with extremely tight power budgets, the scheme can still quickly initiate communication, fully demonstrating its adaptability to weak signal scenarios. Furthermore, this invention offers a wider signal acquisition range. Addressing the tens to hundreds of kilohertz Doppler frequencies commonly found in low-Earth orbit satellite links, this invention expands the integer frequency offset hypothesis set, pre-generating multiple local references to participate in parallel matching, with configurable search step size and search boundaries. Experimental results demonstrate that the scheme can cover frequency drift ranges exceeding two megahertz without increasing the number of symbols, significantly exceeding the upper limit of most current synchronization algorithms, which can only tolerate single subcarrier spacing deviations. This "large step size full coverage" characteristic enables the terminal to maintain rapid initial packet acquisition and smooth subsequent tracking even at extremely low elevation angles and rapid scene switching.

[0074] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of the present invention is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but all such changes and modifications fall within the scope of protection of the present invention.

Claims

1. A time-frequency synchronization method, characterized in that, include: The received signal is acquired, and iterative equivalent matched filtering is performed on the received signal based on the symbol delay to obtain the preliminary location result of the burst signal; Local training symbols are generated based on a preset frequency domain spread spectrum sequence. The local training symbols and the preliminary location results of the burst signal are used for parallel matched filtering to obtain joint synchronization detection results.

2. The time-frequency synchronization method according to claim 1, characterized in that, Iterative equivalent matched filtering of the received signal based on symbol delay includes, Construct a pair of dual time windows with symbol delay; use the dual time windows to perform equivalent matched filtering on the received signal by sliding it along the time axis with a preset step size to obtain the preliminary location result of the burst signal.

3. A time-frequency synchronization method according to claim 1 or 2, characterized in that, The symbol delay is the delay of two time windows, which is equal to an integer multiple of the complete symbol period in the received signal.

4. The time-frequency synchronization method according to claim 3, characterized in that, The preset step size includes a first step size and a second step size; a coarse search is performed in the time domain with the first step size to quickly locate the time segment where a matching peak may exist; when the peak is detected, a fine search is performed in the neighborhood with the second step size to accurately determine the starting position of the burst signal.

5. The time-frequency synchronization method according to claim 4, characterized in that, The first step length does not exceed the length of the burst signal pilot repeat structure, and the second step length is the length of a single symbol; or the first step length does not exceed the length of the burst signal pilot repeat structure, and the second step length is a gradual shortening within the range of a single symbol length.

6. A time-frequency synchronization method according to claim 4 or 5, characterized in that, Local training symbols are generated based on a pre-set frequency domain spreading sequence, including: The frequency domain spread spectrum sequence is cyclically shifted in the frequency domain according to different assumed subcarrier offsets, and multiple sets of offset reference sequences are obtained after fast Fourier transform. Each set of bias reference sequences is converted into local training symbols corresponding to the time domain.

7. The time-frequency synchronization method according to claim 6, characterized in that, The same reference symbols as the frequency domain spreading code of the transmitter are used as the frequency domain spreading sequence.

8. The time-frequency synchronization method according to claim 1, characterized in that, Parallel matched filtering is performed using local training symbols and preliminary burst signal localization results to obtain joint synchronization detection results, including... The local training symbols are correlated with the preliminary localization results of the burst signals in parallel to obtain the correlation value sequence under the corresponding bias assumption. Matching filters for all integer offset values ​​within a preset frequency offset search range are determined by correlation peak detection; and an integer multiple carrier frequency offset estimate matching the received signal is determined based on the global maximum correlation peak.

9. The time-frequency synchronization method according to claim 8, characterized in that, Integer multiple carrier frequency offset estimation is determined by the subcarrier offset and subcarrier spacing frequency corresponding to the global maximum correlation peak.

10. A time-frequency synchronization method according to claim 9, characterized in that, The fractional frequency offset estimate is determined based on the integer multiple carrier frequency offset estimate and the phase change of the global maximum correlation peak.

11. The time-frequency synchronization method according to claim 10, characterized in that, The starting position of the synchronous training symbols is determined by the time window where the global maximum correlation peak is located.

12. A time-frequency synchronization system, characterized in that, The system includes, The burst signal localization module is used to acquire the received signal and perform iterative equivalent matched filtering on the received signal based on the symbol delay to obtain the preliminary localization result of the burst signal; The joint synchronization module is used to generate local training symbols based on a preset frequency domain spread spectrum sequence, and to perform parallel matched filtering using the local training symbols and the preliminary location results of the burst signal to obtain the joint synchronization detection results.

13. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 11.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 11.

Images