Frequency offset estimation method and device for large frequency offset scenario

Abstract

This invention provides a frequency offset estimation method and apparatus for scenarios with large frequency offset. The method includes: performing windowing processing on the delay conjugate multiplication sequence of a signal sequence under different target delay values ​​based on a preset detection time window, obtaining multiple window data corresponding to the delay conjugate multiplication sequence; summing the window data to obtain a summation result corresponding to each window data; determining the optimal frequency offset phase metric corresponding to different target delay values ​​based on the summation result; and merging the optimal frequency offset phase metrics corresponding to each target delay value to obtain a frequency offset estimate of the signal sequence. This invention effectively improves the accuracy of carrier frequency offset estimation.

Description

Technical Field

[0001] This invention relates to the field of wireless local area network baseband signal technology, and in particular to a frequency offset estimation method and apparatus for scenarios with large frequency offset. Background Technology

[0002] When existing 802.11b receivers perform frequency offset estimation before the pre-preamble detection stage, there are many problems, the most important of which are noise sensitivity and signal-to-noise ratio (SNR) threshold issues.

[0003] In standard Direct Sequence Spread Spectrum (DSSS) receivers, pseudo-code synchronization must be achieved before despreading can begin. However, the despreading process is extremely sensitive to frequency offsets. The traditional method is to "acquire the code first, then estimate the frequency"; but under large frequency offsets (such as in high dynamic Doppler environments), code acquisition cannot be completed without prior frequency compensation. If a large frequency offset exists, the correlation peak during despreading drops sharply. When the frequency offset exceeds a certain range (typically 1 / 2x, where x is the correlation integration time), the correlation peak may even be completely submerged in noise, making it impossible to acquire the pseudo-code.

[0004] Most existing solutions, such as algorithms based on delay autocorrelation, perform poorly in low signal-to-noise ratio (SNR) environments. Directly multiplying the received signal with its delayed version using autocorrelation algorithms without correction amplifies additive white Gaussian noise, causing a sharp increase in the variance of the estimate and a drastic deterioration in estimation performance as the SNR worsens. In edge coverage areas, large frequency offset estimation errors prevent the compensated signal from triggering the preamble detector, leading to system synchronization loss and severely impacting receiver operation. DSSS systems typically operate at levels far below ambient noise, making the extraction of large frequency offset information in strong noise extremely challenging. Furthermore, using traditional "blind frequency scanning" methods for frequency offset estimation before synchronization results in an excessively large search space and excessively long access delay (Time-to-First-Fix).

[0005] Therefore, there is an urgent need for a frequency offset estimation method and device for scenarios with large frequency offset to solve the above problems. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a frequency offset estimation method and apparatus for scenarios with large frequency offset.

[0007] This invention provides a frequency offset estimation method for scenarios with large frequency offset, including: Based on a preset detection time window, the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​is windowed to obtain multiple windowed data corresponding to the delay conjugate multiplication sequence. The data from each window are summed to obtain the summation result corresponding to each window data. Based on the summation result, the optimal frequency offset phase metric corresponding to different target delay values ​​is determined; The optimal frequency offset phase metric corresponding to each of the target delay values ​​is combined and calculated to obtain the frequency offset estimate of the signal sequence.

[0008] According to the present invention, a frequency offset estimation method for large frequency offset scenarios is provided, the method further includes: Based on a preset set of delay values, a delay conjugate multiplication operation is performed on the signal sequence to obtain the delay conjugate multiplication sequence corresponding to different target delay values; Each delay value in the preset delay value set corresponds to a correction code.

[0009] According to the frequency offset estimation method for large frequency offset scenarios provided by the present invention, the step of accumulating and summing the data of each window to obtain the accumulating and summing result corresponding to each window data includes: Obtain the target correction code and the corresponding index set; wherein, the index set represents the set of positions of each group to be calculated in the correction code; Based on the target correction code and the index set, the data of each window is accumulated and summed to obtain the accumulated summation result corresponding to each window data.

[0010] According to the frequency offset estimation method for large frequency offset scenarios provided by the present invention, the step of determining the optimal frequency offset phase metric corresponding to different target delay values ​​based on the accumulated summation result includes: Based on the summation result of the largest modulus among all the windowed data of each of the delay conjugate multiplication sequences, the optimal frequency offset phase metric corresponding to different target delay values ​​is determined.

[0011] According to the present invention, a frequency offset estimation method for large frequency offset scenarios is provided, the method further includes: The summation result is cached based on a preset cache duration, wherein the preset cache duration is determined based on the length of the preset detection time window.

[0012] According to the frequency offset estimation method for large frequency offset scenarios provided by the present invention, the step of merging and calculating the optimal frequency offset phase metric corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence includes: The weighted phase term is calculated based on the magnitude, phase angle and target delay value corresponding to each of the optimal frequency offset phase metrics; The weighted phase terms corresponding to all the target delay values ​​are summed to obtain the merged phase metric values. Based on the combined phase metric values ​​and the baseband sampling frequency, the frequency offset estimate of the signal sequence is calculated.

[0013] The present invention also provides a frequency offset estimation device for large frequency offset scenarios, comprising: The detection window data selection module is used to perform windowing processing on the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​based on a preset detection time window, so as to obtain multiple window data corresponding to the delay conjugate multiplication sequence. The cumulative summation module is used to perform cumulative summation processing on the data of each window to obtain the cumulative summation result corresponding to each window data. The extreme value selection module is used to determine the optimal frequency offset phase metric corresponding to different target delay values ​​based on the summation result. The frequency offset phase angle estimation module is used to merge and calculate the optimal frequency offset phase metric corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the frequency offset estimation method for large frequency offset scenarios as described above.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the frequency offset estimation method for large frequency offset scenarios as described above.

[0016] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the frequency offset estimation method for large frequency offset scenarios as described above.

[0017] The frequency offset estimation and apparatus for large frequency offset scenarios provided by this invention performs delay conjugate multiplication on the signal sequence according to different target delay values ​​through a preset detection time window, and then performs windowing processing to obtain multiple window data; then, the window data are accumulated and summed, and the optimal frequency offset phase metric corresponding to different target delay values ​​is determined based on the results; finally, the optimal frequency offset phase metric of each target delay value is calculated by merging, thereby obtaining the frequency offset estimate of the signal sequence, effectively improving the accuracy of carrier frequency offset estimation. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the frequency offset estimation method for large frequency offset scenarios provided by the present invention. Figure 2 This is a schematic diagram of the structure of the preset detection time window provided by the present invention; Figure 3 One of the schematic diagrams of the correction code provided by the present invention; Figure 4 A second schematic diagram of the correction code provided by the present invention; Figure 5 The third schematic diagram of the correction code provided by the present invention; Figure 6 The fourth schematic diagram of the correction code provided by the present invention; Figure 7 This is a schematic diagram of the overall architecture of the frequency offset estimation method for large frequency offset scenarios provided by the present invention. Figure 8 This is a schematic diagram of the overall process of the frequency offset estimation method for large frequency offset scenarios provided by the present invention. Figure 9 This is a schematic diagram of the simulation results provided by the present invention; Figure 10 This is a schematic diagram of the frequency offset estimation device for large frequency offset scenarios provided by the present invention. Figure 11 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0021] Although high-speed protocols such as 802.11ax (Wi-Fi 6) exist, 802.11b remains the protocol that Wi-Fi chips must maintain backward compatibility with. It also has the following characteristics: Channel characteristics: The 2.4GHz band is subject to severe interference and significant multipath effects.

[0022] Hardware constraints: Early hardware computing power was limited, requiring synchronization algorithms to be completed quickly under low signal-to-noise ratio (SNR).

[0023] The core problem is that imperfections in the crystal oscillator cause a frequency offset (CFO) between the transmitter and receiver. If the CFO cannot be resolved during the preamble detection stage, the receiver's despreading efficiency will decrease significantly.

[0024] In wireless communication, especially in receiver design based on IEEE 802.11b and DSSS technology, carrier frequency offset is one of the core challenges affecting system performance. Estimating the frequency offset before or simultaneously with preamble detection is crucial for ensuring subsequent demodulation, despreading, and data recovery.

[0025] In the IEEE 802.11b (DSSS / CCK) system, frequency offset estimation (FOE) is crucial before or during the initial stage of preamble detection because a large frequency offset can prevent the Barker code correlator from "locking" onto the signal.

[0026] Because 802.11b uses direct sequence spread spectrum, the preamble consists of a synchronization field modulated by an 11-chip Barker sequence (128 scrambled bits "1" in long format and 56 scrambled bits "1" in short format). Receivers can estimate the frequency shift by measuring the phase rotation of these known Barker-coded bits over time. Most related techniques focus on two main types: differential detection (frequency shift invariant) and incoherent energy detection before fine estimation.

[0027] The development of carrier synchronization technology has evolved from analog to digital, and from open-loop to closed-loop: In early communication systems, frequency offset correction was mainly achieved using analog phase-locked loops (Analog PLLs) at the radio frequency end. Because this technology was applied to the analog front-end of the receiver and used analog signals instead of digital signals, its acquisition speed was slow, its accuracy and precision were relatively low, and it was difficult to handle sudden frequency offsets caused by rapid fading.

[0028] With the development of digital circuits, mid-stage communication systems adopted a coarse synchronization based on time-domain autocorrelation, utilizing the repetition of the synchronization (SYNC) field in the 802.11b preamble for digitization. This technique introduced a delay-and-correlate algorithm, which does not require prior knowledge of the data content, but only utilizes the periodicity of the signal. Since the periodicity of the 802.11b preamble sequence can only be obtained after synchronization, before that, only pseudo-periodic characteristics exist; that is, the possible phase reversals between preamble bits are uncertain in their occurrence, but the period size is known.

[0029] The current 802.11b receiver immediately enters parallel processing mode after detecting energy triggering, and uses the Coordinate Rotation Digital Computer (CORDIC) algorithm to perform phase rotation calculation in a very short time; and combines it with multi-path sampling to suppress multipath interference while compensating for frequency offset.

[0030] Prior to 802.11b Preamble Detection, the most mainstream and closest existing technical solutions for frequency offset estimation typically include the following three: 1. Variants of the Schmidl & Cox algorithm: Although the S&C algorithm was originally designed for Orthogonal Frequency Division Multiplexing (OFDM), its core idea—using the repetition of training sequences for autocorrelation—has been widely adopted in the DSSS scheme.

[0031] The algorithm works by multiplying the received Barker code sequence by its conjugate multiplication with a version delayed by 1 microsecond (one symbol period). However, this algorithm does not fully utilize the characteristics of the Barker code sequence itself and is very sensitive to noise, resulting in insufficient accuracy at low signal-to-noise ratios.

[0032] 2. Differential Phase Estimation: As the most commonly used technique in 802.11b, its principle is that the 802.11b preamble, after being processed by scrambling, is a sequence of all 1s, which, after spreading, exhibits extremely strong autocorrelation. The receiver obtains the peak value through a correlator, calculates the phase rotation between adjacent peak values, and thus infers the frequency deviation. However, this technique requires frequency deviation estimation to be performed when synchronization is complete, and generally requires the sequence to be detected to be under oversampling preconditions, as well as a very long observation period.

[0033] 3. Data-Aided Loop: This technique utilizes a known preamble sequence (128 bits of the SYNC field) as a local reference while detecting the preamble. The principle is to perform matched filtering of the received signal against the locally known Barker code, and then extract the frequency offset using an error discriminator. As the most accurate method currently available, it typically follows "coarse frequency offset estimation" to fine-tune the frequency and ensure demodulation of subsequent payloads. However, this technique is complex, requires auxiliary data, and is difficult to perform before preamble detection.

[0034] In 802.11b receiver design, although existing technologies are very mature, several challenging technical bottlenecks still exist when estimating frequency offset before the pre-preamble detection stage: I. Noise sensitivity and signal-to-noise ratio (SNR) threshold issues: Most existing solutions (such as algorithms based on delay autocorrelation) perform extremely poorly in low signal-to-noise ratio environments.

[0035] Multiplicative noise amplification: Existing autocorrelation algorithms typically multiply the received signal by its delayed version conjugate. This operation, while extracting the phase difference, also amplifies the additive white Gaussian noise (AWGN), leading to a sharp increase in the variance of the estimator and causing the estimation performance to deteriorate drastically as the signal-to-noise ratio worsens.

[0036] Detection failure: In the edge coverage area (when the signal is weak), due to the excessive frequency offset estimation error, the compensated signal still cannot trigger the preamble detector, causing the system to lose synchronization.

[0037] II. The contradiction between estimation range and accuracy: Before preamble detection, the receiver often doesn't know the exact symbol timing, and within the finite detection time, this also means it's unclear which segment of the entire preamble the detected signal belongs to. Specifically, in 802.11b or similar DSSS systems, when using delay multiplication for frequency offset estimation, a single delay often faces a trade-off between estimation range and estimation accuracy, leading to the following issues in existing technologies: Short delay: Large estimation range (phase is unlikely to exceed ±π), but sensitive to noise and low accuracy. If a shorter delay interval is used for autocorrelation, the estimation range is large (covering more than ±100kHz), but the noise resistance is poor and the accuracy is low.

[0038] Long Delay: High accuracy (noise is averaged), but small range, prone to phase ambiguity. If the full Barker code period (1μs) is used for estimation, the accuracy is high, but the maximum frequency offset range that it can capture is severely limited due to the phase entanglement effect.

[0039] Existing technologies often require a complex two-stage structure of "coarse estimation + fine estimation", which increases hardware overhead and processing latency.

[0040] III. The conflict between computing resources and real-time performance: In order to complete the estimation before preamble detection, the algorithm must produce results in an extremely short time (typically on the order of microseconds).

[0041] If detection is required at a sampling rate higher than the Nyquist rate, the pre-processing stage needs to be designed as an oversampling system, adding sampling rate adaptation processing, which will increase system resource overhead.

[0042] However, many existing solutions rely on high-performance Fast Fourier Transform (FFT) or large-point CORDIC rotation. In chip applications like 802.11b, which require low power consumption and low cost, these algorithms consume a large number of logic gates and power.

[0043] If frequency offset estimation takes too long, it can cause the backend Barker code correlator to miss the optimal synchronization window, resulting in a halt in the system processing pipeline.

[0044] Fourth, it heavily relies on "pseudo-periodicity" and is susceptible to multipath interference: While the 802.11b preamble is repetitive, this periodicity can be disrupted in complex indoor multipath environments. Specifically, this manifests in the following two ways: Inter-symbol interference: When processing multipath reflected signals, existing technologies cannot distinguish between "phase deflection caused by frequency offset" and "phase deflection caused by multipath interference" because the phase shift caused by the reflected waves is random and overlapping.

[0045] False alarm problem: In a strong interference environment, existing technology may misjudge some interference signals as 802.11b signals with frequency offset characteristics, thereby erroneously triggering subsequent processing circuits.

[0046] In summary, accurate frequency offset estimation requires accurate symbol timing (knowing the start of each chip), and accurate symbol timing (preamble detection) depends on pre-corrected frequency offset. Most existing schemes use "blind estimation" or "sliding window averaging," but these methods often cannot keep up with the ebb of the preamble SYNC field when the signal is rapidly fading or there is narrowband interference.

[0047] An advanced 802.11b frequency offset estimation scheme prior to preamble detection needs to meet the following specifications: Capture Range: Whether it can cover frequency deviations above ±100kHz; Convergence speed: Whether the estimation can be completed within the first 20-30 microseconds of the SYNC field; Computational resources: Is a large number of multipliers required, or only simple CORDIC shift operations?

[0048] In 802.11b, although the SYNC field is generated by scrambling all "1"s, after differential binary phase shift keying (DBPSK) modulation, there may indeed be a 180° polarity flip between adjacent 11-chip Barker code groups (i.e., a preamble bit).

[0049] If long-delay accumulation (across bit boundaries) is performed directly, the phase rotation will jump 180° due to polarity reversal, causing the frequency offset estimation to completely fail or produce huge errors. Even if there is no polarity reversal between bits, the same situation will occur if there is a large time interval between bits in the presence of a large frequency offset. In order to improve the estimation performance under low signal-to-noise ratio conditions, it is necessary to increase the observation accumulation length.

[0050] In existing methods, the squaring operation of the sequence is used to overcome the bit phase reversal operation, but this amplifies the effect of noise, especially under low signal-to-noise ratio conditions. The squaring operation of the sequence causes the noise variance characteristics to change very much, resulting in a sharp deterioration in the frequency offset estimation performance, or even making it unusable.

[0051] Based on the characteristics of the 802.11b spread spectrum Barker code and the phase distribution characteristics of multiple preamble bits spanning multiple bits, this invention appropriately "corrects" the sequence after "delayed conjugate multiplication," and then performs cumulative observations on the corrected sequence to reduce the influence of noise, thereby accurately capturing the frequency offset phase difference information introduced by the carrier frequency offset. Moreover, this frequency offset phase difference information does not depend on the system's oversampling preconditions, making it more widely applicable. Furthermore, this invention can perform appropriate observation accumulation as needed, ensuring the ability to observe symbol timing synchronization; that is, while dynamically estimating the frequency offset, it also dynamically estimates the timing synchronization, making timing synchronization and frequency offset estimation mutually dependent. Finally, based on the principle of the most significant accumulated energy, the estimate at the most reliable position is used as the final frequency offset estimate within the observation window time period.

[0052] Furthermore, existing methods essentially utilize the property that "each bit of the original bit sequence is extended by a finite-length Barker code" during Barker code spreading, meaning that the Barker code appears periodically (but may collectively reverse due to the polarity of the original bits). Therefore, when considering "delay conjugate multiplication," only the delay of consistent Barker code length is used, which limits the maximum frequency offset estimation range to only the reciprocal of half the length of the Barker code sequence.

[0053] This invention, however, does not have this limitation and can support arbitrary delay lengths. Theoretically, its maximum frequency offset estimation range can reach the reciprocal of half the time of the Barker code chip length. Furthermore, because this invention can utilize more Barker code delay characteristics, it can also merge the frequency offset phase estimates corresponding to different "delay conjugate multiplications," further improving the performance of frequency offset estimation. Existing technologies only utilize the delay of the Barker code sequence length, making it impossible to perform the aforementioned frequency offset phase estimate merging and thus unable to achieve performance improvement in frequency offset estimation.

[0054] Because this invention introduces a "correction" after the delayed conjugate multiplication, it greatly expands the range of accumulated observations. Theoretically, it can be applied to frequency offset estimation for longer observation periods, rather than being limited to a duration of only 2 to 3 bits. This is beneficial for improving the performance of frequency offset estimation under low signal-to-noise ratio conditions. Existing technologies require prior knowledge of the exact starting position of the original bit sequence, and then use the phase accumulation of the delayed conjugate multiplication to estimate the frequency offset. This is essentially frequency offset estimation performed after time synchronization. However, this invention adopts a method of "significantly accumulating energy" within a dynamic observation window during the synchronization estimation process. Based on approximate synchronization, the carrier frequency offset within the observation window can be estimated. That is, within the observation window, time synchronization is "synchronously" tested and the carrier frequency offset is estimated. This greatly reduces the overall processing delay of frequency offset estimation and is applicable to more applications.

[0055] Furthermore, most of the computational parts in this invention can be processed in parallel. For example, a series of processes for different selected delay values, including delay multiplication to search for the best estimate; and in the corrected cumulative summation, traversal within the observation window time. Some operations can also be processed in parallel, which can significantly reduce the processing latency.

[0056] The technology described in this invention improves the frequency offset detection sensitivity near the noise floor by modifying the accumulation structure. During the frequency offset pre-estimation process before DSSS synchronization, this technology first narrows the search range by an order of magnitude (not exceeding the chip rate), and then significantly improves the system's response speed by simultaneously searching for possible preamble sequences and estimating the frequency offset. Furthermore, this method does not require the sequence to be processed to be under oversampling conditions, greatly reducing the complexity of the system's pre-processing.

[0057] Figure 1 The flowchart of the frequency offset estimation method for large frequency offset scenarios provided by the present invention is shown below. Figure 1 As shown, this invention provides a frequency offset estimation method for scenarios with large frequency offset, including: Step 101: Based on a preset detection time window, perform windowing processing on the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​to obtain multiple windowed data corresponding to the delay conjugate multiplication sequence.

[0058] In this invention, a "delayed conjugate multiplication" operation is performed on the input signal sequence. Specifically, according to different target delay values, each element in the signal sequence is multiplied by the element after the corresponding delay and conjugate operation, thereby obtaining a delayed conjugate multiplication sequence under different target delay values.

[0059] Next, arbitrarily select a time point from the signal sequence as the starting point, and continuously select a certain number of data points (e.g., M bit durations) as a detection time window according to the preset detection time window setting. Furthermore, the start time of the next detection time window must be at least a certain interval (not less than M). overlap The data (within a few bits of duration) is used to avoid excessive data overlap. The delay conjugate multiplication sequence for each target delay value is windowed in the same way, resulting in multiple windowed data corresponding to each delay conjugate multiplication sequence.

[0060] Step 102: Perform cumulative summation on the data of each window to obtain the cumulative summation result corresponding to each window data.

[0061] In this invention, the data in each window under each target delay value are summed. During the processing, the operation is performed based on the "correction code" (bipolar binary symbol sequence) corresponding to the target delay value.

[0062] Specifically, the windowed data is grouped according to a certain grouping method, and then the data in each group is multiplied by the correction code before being summed. Simultaneously, for some special subscript sets in the correction code (i.e., subscript sets K1 and K2, used to represent the signal positions to be differentiated in the correction code), additional addition and subtraction operations are required based on the different values ​​(+1 or -1) of the grouping weighting coefficients. Although the value conditions require iterative calculation, the use of grouped summation greatly reduces the number of addition operations, improving processing efficiency. Finally, the summation result corresponding to each window of data is obtained.

[0063] Step 103: Based on the summation result, determine the optimal frequency offset phase metric corresponding to different target delay values.

[0064] To improve the frequency offset estimation performance at low signal-to-noise ratios (SNR), this invention caches the summation results corresponding to different target delay values, thereby improving diversity gain. Furthermore, to make the frequency offset estimation insensitive to the start position of the preamble, the caching of these data will continue for a period of time (selectably M-1 bit times, where M is the length of a preset detection time window).

[0065] Then, given the target delay value, the series of accumulated summation results obtained from different combinations of cached values ​​are selected according to the principle of the largest modulus. That is, under the same target delay value, the magnitudes of different accumulated summation results are compared, and the start time of the sequence corresponding to the result with the largest modulus is selected. This result is considered to be the most likely frequency offset phase measure under the target delay value, that is, the optimal frequency offset phase measure.

[0066] Step 104: Combine and calculate the optimal frequency offset phase metric corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence.

[0067] In this invention, after obtaining the optimal frequency offset phase metric corresponding to each target delay value, since there are multiple target delay values ​​(d1, ..., d...), D For each target delay, the optimal frequency offset phase metric is weighted according to its magnitude. The weighting coefficients are related to the target delay. The weighted results are then combined to obtain a combined frequency offset phase angle estimate. Finally, based on this frequency offset phase angle estimate and the baseband sampling frequency, the frequency offset estimate of the signal sequence is obtained through a conversion relationship.

[0068] The frequency offset estimation method for large frequency offset scenarios provided by this invention performs delay conjugate multiplication on the signal sequence according to different target delay values ​​through a preset detection time window, and then performs windowing processing to obtain multiple window data; then, the window data are accumulated and summed, and the optimal frequency offset phase metric corresponding to different target delay values ​​is determined based on the results; finally, the optimal frequency offset phase metric for each target delay value is calculated by merging, thereby obtaining the frequency offset estimate of the signal sequence, which effectively improves the accuracy of carrier frequency offset estimation.

[0069] Based on the above embodiments, the method further includes: Based on a preset set of delay values, a delay conjugate multiplication operation is performed on the signal sequence to obtain the delay conjugate multiplication sequence corresponding to different target delay values; Each delay value in the preset delay value set corresponds to a correction code.

[0070] In this invention, the input... The sequence, starting from an arbitrarily chosen time, uses a continuous M-bit duration as a detection window (i.e., a preset detection time window), and the start of the next detection window is at least M bits longer than the given time. overlap The duration of each bit.

[0071] Figure 2 This is a schematic diagram of the structure of the preset detection time window provided by the present invention, which can be referred to. Figure 2 As shown, (d=8, M=4, M) overlap =1) As an example, the data within each detection window time will output an estimated frequency shift, denoted as the data output of the p-th detection window. In the case of no oversampling, the Barker code length is 11; in the case of 2x oversampling, the Barker code length is 22, and so on. This invention is described with reference to the case of no oversampling; the oversampling case still follows the method described in this invention.

[0072] Furthermore, to comprehensively analyze the characteristics of the signal, especially to detect frequency shifts, a set of different delay values ​​is preset, forming a pre-defined delay value set. For example, the pre-defined delay value set might include d=1, 2, 3, etc. Different delay values, such as 11, are used. The choice of these delay values ​​is usually related to the characteristics of the signal, processing requirements, and the design of subsequent algorithms.

[0073] For the input signal sequence For each target delay value d in the preset delay value set, a delay conjugate multiplication operation must be performed. The mathematical expression for this operation is: ; in, express The conjugate of . By delaying the conjugate multiplication operation, a new sequence can be obtained. The delayed conjugate multiplication sequence reflects the correlation information of signals under different delay conditions.

[0074] Taking d=3 as an example, for the input signal sequence ,calculate The delayed conjugate multiplication sequence corresponding to a delay value of 3 is obtained. Similarly, other delay values ​​d can be calculated in the same way. sequence.

[0075] Since the preset delay value set contains multiple different delay values, performing the aforementioned delay conjugate multiplication operation on each delay value will yield a sequence of delay conjugate multiplications corresponding to different target delay values. For example, when the preset delay value set is {1, 2, 3}, the operation will result in... , and These are three different delayed conjugate multiplication sequences. These sequences contain characteristic information of the signal under different delay conditions, providing basic data for subsequent signal analysis and processing.

[0076] Correction code It is a bipolar binary symbol sequence associated with a specific delay value d. In subsequent signal processing, the correction code is used to modify the delay conjugate multiplication sequence. Further adjustments and optimizations will be made to extract information such as signal frequency offset more accurately.

[0077] In this invention, for each delay value d in the preset delay value set, there exists a corresponding correction code. The specific correspondence is as follows: When d=1, =[-1 -1 1 -1 -1 1 1 -1 1 1], k=0, 1, , 9, K1= 10, K2={}; When d=2, =[ 1 -1 -1 1 -1 1 -1 -1 1], k=0, 1, ..., 8, K1={9}, K2={10}; When d=3 =[1 1 1 1 -1 -1 -1 -1], k=0, 1, ..., 7; K1={8, 10}, K2={9}; When d=4 =[-1 -1 1 1 1 -1 -1], k=0, 1, ..., 6; K1={7, 8}, K2={9, 10}; When d=5 =[1 -1 1 -1 1 -1], k=0, 1, ..., 5; K1={6, 10}, K2={7, 8, 9}; When d=6 =[1 -1 -1 -1 1], k=0, 1, ..., 4; K1={5, 7, 9}, K2={6, 8, 10}; When d=7 =[1 1 -1 -1], k=0, 1,...,3; K1={4, 5, 9, 10}, K2={6, 7, 8}; When d=8 =[-1 1 -1], k=0, 1, 2; K1={3, 4, 5, 6}, K2={7, 8, 9, 10}; When d=9 =[-1 1], k=0, 1; K1={2, 5, 7, 10}, K2={3, 4, 6, 8, 9}; When d=10 =[-1], k=0, 1; K1={1, 2, 4, 5, 8}, K2={3, 6, 7, 9}; When d=11 = [], K1={0~10}, K2={}.

[0078] Where K1 and K2 represent the sets of positions of each group to be calculated in the correction code. Figure 3 One of the schematic diagrams of the correction code provided by the present invention. Figure 4 The second schematic diagram is of the correction code provided by the present invention. Figure 5 The third schematic diagram of the correction code provided by this invention. Figure 6 The fourth schematic diagram of the correction code provided by the present invention, wherein when d=3, the correction code can be referred to Figure 3 As shown; when d=5, the correction code can be referenced. Figure 4 As shown; when d=6, the correction code can be referenced. Figure 5 As shown; when d=9, the correction code can be referenced. Figure 6 As shown.

[0079] In this invention, different delay values ​​correspond to correction codes of different lengths and values. These correction codes are determined based on specific requirements and theoretical derivations when designing signal processing algorithms. They can help the algorithms better adapt to signal characteristics under different delay conditions, collect more frequency offset phase information, and thus improve the accuracy and reliability of frequency offset estimation.

[0080] Based on the above embodiments, the step of performing cumulative summation on the data of each window to obtain the cumulative summation result corresponding to each window data includes: Obtain the target correction code and the corresponding index set; wherein, the index set represents the set of positions of each group to be calculated in the correction code; Based on the target correction code and the index set, the data of each window is accumulated and summed to obtain the accumulated summation result corresponding to each window data.

[0081] In this invention, a multi-bit delay-related pattern module can be set in the signal processing flow. This module generates a corresponding correction code based on a preset delay factor d (i.e., delay value). This is the target correction code. This correction code is a bipolar binary symbol sequence whose length and value vary depending on the delay value d. For the specific structure, please refer to the above embodiment.

[0082] The index sets include K1 and K2, associated with a specific delay value d. The multi-bit delay-related pattern module generates the correction code. Simultaneously, the corresponding sets K1 and K2 are determined. For example, when d=3, K1={8, 10}, K2={9}; when d=5, K1={6, 10}, K2={7, 8, 9}. These subscript sets represent the sets of positions in the modified code to be grouped for calculation, that is, which positions' data need to be calculated according to specific rules (and grouping weighting coefficients) during subsequent accumulation and summation operations. (After multiplication, addition and subtraction are performed for processing.)

[0083] In this invention, the delayed conjugate multiplication sequence The process involves taking an arbitrary time starting from which M consecutive bits are used as a detection time window, and then outputting the windowed data. ,in, Windowed data refers to signal data selected within different detection time windows.

[0084] In this invention, summation can be performed using the following summation formula: ; Part One: It is to divide the data into windows. With correction code The corresponding parts are multiplied and then accumulated. A correction code is used to adjust the signal data to highlight components related to features such as frequency shift. For example, the positive and negative values ​​in the correction code can be designed according to the characteristics of the signal, thereby enhancing useful signal components and suppressing interference components during the accumulation process.

[0085] Part Two: ; This part involves the grouping weighting coefficient. Its value is either +1 or -1. The element k in sets K1 and K2 corresponds to the position in the signal data that requires special processing. For k in set K1, the corresponding signal data is... After multiplication, the data is accumulated; for k in set K2, the corresponding signal data is... After multiplying, subtract from the sum. This grouped differential processing method can further optimize signal characteristics and improve the accuracy of frequency offset estimation.

[0086] because The value of is unknown, so a traversal approach is needed to calculate it. Although the traversal requires a maximum of 2... M-2 The complete calculation takes only 2 steps, but by using grouping and summing, it only requires 2 steps. M-1 +11M additions. Specifically, it can perform 11 summations within each bit (by traversing...). Parallel processing is used, and only 2 are used subsequently. M-1 All processing can be completed in one serial addition. This optimization method improves computational efficiency, enabling the rapid acquisition of the cumulative summation results corresponding to each window of data in real-time signal processing.

[0087] Through the above summation operation, for each detection time window (such as the first...), p Window data within each detection time window Each of these will result in a corresponding summation result. These results can be used for subsequent signal processing tasks such as frequency offset estimation, providing important data support for accurately analyzing signal characteristics. For example, in frequency offset estimation, these accumulated summation results can be used for further calculations and analysis to obtain a more accurate frequency offset estimate.

[0088] Based on the above embodiments, determining the optimal frequency offset phase metric corresponding to different target delay values ​​based on the accumulated summation result includes: Based on the summation result of the largest modulus among all the windowed data of each of the delay conjugate multiplication sequences, the optimal frequency offset phase metric corresponding to different target delay values ​​is determined.

[0089] In this invention, a correction code is generated by a multi-bit delay correlation pattern module, and after performing a series of operations with the windowed data, an accumulated summation result is obtained. Different delay values ​​d will generate different correction codes and operation results. The sequences obtained by processing different delay values ​​d are delayed conjugate multiplication sequences.

[0090] When estimating the frequency offset of a signal, a metric is needed to measure the accuracy of the estimation. The optimal frequency offset phase metric is the value that most accurately reflects the true frequency offset within this metric system. In one embodiment, for each delay value d, the grouping weighted coefficients are iterated through... A series of results And cache it. This cached data contains different values ​​at a specific latency value d. The combination of the corresponding summation results provides a foundation for finding the optimal frequency offset phase metric in the future.

[0091] For each latency value d, from a series of cached values... In the middle, selection is based on the largest modulus value, that is, for each n and Combinations, calculate Then find the largest value among them. n and The formula for selecting the extreme value of a combination is: in, Indicates the first p The first detection time window within the first detection time window At each sampling time point, for a given delay value d, the summation result is... The magnitude reaches its maximum. That is, under the condition of delay value d, the signal characteristics corresponding to this moment can most accurately reflect the frequency offset information, and it is the best frequency offset phase measure.

[0092] Since each delay value d corresponds to a series of summation results, through the above extreme value selection process, each delay value d obtains an optimal frequency offset phase metric. The optimal frequency offset phase measure obtained under different delay values ​​reflects the frequency offset characteristics of the signal under different delay conditions, providing a basis for subsequent comprehensive analysis of frequency offset information. By comparing the optimal frequency offset phase measures under different delays, the frequency offset characteristics of the signal under different delays can be further analyzed, providing more comprehensive information for accurate frequency offset estimation.

[0093] Based on the above embodiments, the method further includes: The summation result is cached based on a preset cache duration, wherein the preset cache duration is determined based on the length of the preset detection time window.

[0094] In this invention, a preset detection time window is used to select data within a specific time period from the input signal for analysis. In signal processing, to accurately estimate signal parameters (such as frequency offset), it is necessary to select a representative segment of signal data. The length of the detection time window affects the extraction and analysis of signal features. For example, if the detection time window is too short, it may not contain enough information to accurately estimate the frequency offset; if the detection time window is too long, it may introduce excessive noise and interference, reducing the accuracy of the estimation.

[0095] The length of the detection time window is usually determined based on the characteristics of the signal and the processing requirements. For example, in preamble-based signal detection, considering the structure and characteristics of the preamble, a detection time window is selected for a duration of M consecutive bits. Here, M is determined according to the design of the preamble and the requirements of the signal processing algorithm, reflecting a basic unit of the signal in the time dimension.

[0096] Furthermore, in order to improve the frequency offset estimation performance at low signal-to-noise ratios (SNR), it is necessary to convert the frequency offset estimation performance corresponding to different delay values ​​d. The calculated cumulative sum result All data is fed into the data buffer, thereby improving diversity gain. Diversity gain is a technique that improves system performance by utilizing multiple independent fading signal paths. In frequency offset estimation, the summation results corresponding to different delays can be regarded as signals from different paths. Buffering them allows for the comprehensive utilization of this information, improving the accuracy of the estimation.

[0097] To ensure that frequency offset estimation is insensitive to the start position of the preamble bit, the data needs to be cached for a sustained period. In this invention, a preset cache duration of M-1 bit time is chosen. This is because the length of the detection time window is usually related to M. Choosing M-1 bit time as the cache duration ensures full utilization of the accumulated summation results under different delays while avoiding the introduction of excessive redundant information due to excessively long cache time, and it can adapt to changes in the start position of the preamble bit. For example, if the detection time window is a continuous duration of M bits, then caching M-1 bit time of data can cover sufficient relevant information under different start positions, thereby improving the robustness of frequency offset estimation.

[0098] After determining the preset cache duration, the accumulated summation results are cached according to this duration. Specifically, the accumulated summation results corresponding to different latency values ​​d are cached. The data is sequentially fed into the data cache, which is cached for a preset duration of M-1 bit time. During the caching process, new cumulative summation results are continuously added to the cache, while older data is removed from the cache according to the first-in, first-out (FIFO) principle, ensuring that the cached data is always the cumulative summation result from the most recent M-1 bit time.

[0099] In this invention, when performing frequency offset estimation, the accumulated summation results under different delays can be extracted from the buffer. By analyzing and processing these results, the signal characteristics of different paths and different times are comprehensively considered, thereby more accurately estimating the frequency offset of the signal.

[0100] Based on the above embodiments, the step of merging and calculating the optimal frequency offset phase metric corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence includes: The weighted phase term is calculated based on the magnitude, phase angle and target delay value corresponding to each of the optimal frequency offset phase metrics; The weighted phase terms corresponding to all the target delay values ​​are summed to obtain the merged phase metric values. Based on the combined phase metric values ​​and the baseband sampling frequency, the frequency offset estimate of the signal sequence is calculated.

[0101] In this invention, for each optimal frequency offset phase metric The optimal frequency offset phase metric is calculated to determine its magnitude, reflecting the signal's amplitude, and its phase angle, reflecting the signal's phase information. The optimal frequency offset phase metric is a complex value; its magnitude corresponds to the amplitude of this complex value, and its phase angle corresponds to the angle between this complex value in the complex plane and the positive X-axis. The magnitude and phase angle of the optimal frequency offset phase metric integrate the signal's amplitude and phase information. The amplitude can be seen as a weighting of the phase information; a larger amplitude indicates a potentially greater contribution of the signal at that delay to the frequency offset estimation.

[0102] Target delay value It is the optimal frequency offset phase measurement The corresponding delay values, such as d1, ..., d D Different delay values ​​correspond to the characteristics of the signal under different time delays. In frequency offset estimation, the contribution of signal information under different delays to the final estimation result is different.

[0103] For each target delay value Divide the previously calculated modulus and phase angle by the target delay value. This yields the weighted phase term. By weighting the phase information under different delays, the impact of different delays on frequency offset estimation becomes more reasonable. Because a larger delay value may result in a smaller frequency offset for the same phase change, appropriate adjustments can be made by dividing by the delay value.

[0104] Next, the weighted phase terms corresponding to the D target delay values ​​are summed. This summation operation combines the weighted phase information under different delays to obtain a combined phase metric. This reflects the phase change of the signal over the entire considered delay range, providing more comprehensive and accurate information for subsequent frequency offset estimation. The specific formula for calculating the combined phase metric value is as follows: .

[0105] Baseband sampling frequency Frequency offset is an important parameter in signal processing, representing the frequency at which the baseband signal is sampled. In frequency offset estimation, there is a specific relationship between phase change and frequency offset, and the baseband sampling frequency provides a reference for converting phase change into frequency offset.

[0106] Finally, based on the relationship between phase change and frequency offset, the known phase metric values ​​are combined. Through formula It can calculate the frequency offset estimate of the signal sequence. .

[0107] Figure 7 This is a schematic diagram of the overall architecture of the frequency offset estimation method for large frequency offset scenarios provided by the present invention, which can be referred to. Figure 7 As shown, the input signal It is the starting point of the entire processing flow.

[0108] Input signal First, a delay and multiply operation is performed. During this process, the signal... Multiplying it by its conjugate after delay d, we get Where d is the preset delay value, indicated by the superscript. This indicates the conjugate operation.

[0109] Next, to Perform the data selection process for the detection window. This process starts from... Select data with a duration of M consecutive bits as the detection time window, and the output is... ,in It should be noted that this embodiment assumes the Barker code length is 11 under non-oversampling conditions.

[0110] Meanwhile, the multi-bit delay correlation pattern module provides the corresponding correction code based on the preset delay value d. And the index sets K1(d) and K2(d), where the correction code It is a bipolar binary symbol sequence, and the subscript sets K1(d) and K2(d) are used for subsequent summation operations.

[0111] Furthermore, in the cumulative summation process, based on the correction code... And index sets K1(d) and K2(d), for the detection window data Perform the following calculations: ; in, It is the grouping weighting coefficient, with a value of +1 or -1.

[0112] Then, sum the results. The data is fed into a buffer. The purpose of the buffer is to improve the frequency offset estimation performance at low signal-to-noise ratios (SNR) by utilizing the results corresponding to different delay values ​​d to improve diversity gain. The buffer lasts for M-1 bit time so that the frequency offset estimation is insensitive to the start position of the preamble bits.

[0113] Furthermore, from the cached data, extreme value selection is performed for each latency value d, thereby selecting the summation result with the largest modulus. This serves as the optimal frequency offset phase metric for that delay.

[0114] Finally, frequency offset phase angle estimation is performed based on the optimal frequency offset phase metric corresponding to all delay values ​​d, thereby calculating the combined phase metric value result. Then according to and baseband sampling frequency The frequency offset estimate is calculated. .

[0115] Figure 8 This is a schematic diagram of the overall process of the frequency offset estimation method for large frequency offset scenarios provided by the present invention, which can be referred to. Figure 8 As shown, the specific implementation steps are as follows: 1. Parameter set selection: First, based on the frequency offset estimation requirements, select a set of parameters consisting of different delay values. D ={ d 1,…, d D}, and the maximum allowable processing latency (at least 2 bits for a duration).

[0116] 2. Select a delay value d， This corresponds to a known multi-bit delay-related pattern sequence. , and the subscript sets K1(d) and K2(d).

[0117] 3. Delayed multiplication operation, for each delayed value d , d∊D By utilizing the delay multiplication process, the multiplication of the input sequence with conjugate versions of the input with different delays is calculated, yielding: 4. Corrected cumulative summation: Within the selected detection time window (duration is M bit duration), calculate the current number of bits using the grouping summation method shown in the following formula. p Within each detection window n Undetermined metric value at time: ; in, , .

[0118] 5. In the optimal estimation stage, under a given time delay value d, different... A series of speculations Value, by time All values ​​are cached together, and the value with the maximum energy and the corresponding time are found, i.e.: in, That is to say, in the first p The first detection time window within the first detection time window Each sampling time is determined as the estimated time of the optimal frequency offset phase metric for a given time delay value d.

[0119] 6. Frequency offset phase metric merging calculation, different time delay values d i Below, for the same data segment Repeat steps 2 through 5 above, and then... d i The optimal frequency offset phase metric values ​​are combined as follows: ; The optimal frequency offset estimate is obtained: Through the above process, this invention does not require prior knowledge of the bit start position and bit content information of the preamble-bits. It only needs to use the known "correction" code corresponding to the predetermined "delayed conjugate multiplication" strategy to obtain the frequency offset estimation observation without phase difference bias. By using the "correction" code and three sets of summation methods, the phase difference bias problem caused by the phase difference operation ("delayed conjugate multiplication") due to the phase reversal of the preceding and following bits can be avoided.

[0120] Furthermore, addressing the problem that existing methods for calculating frequency offset phase difference can only accumulate computations over a few identical bit durations without prior knowledge of the preamble bits, this invention, based on a "correction" code and a summation grouping method, is applicable to the accumulation of any number of bit lengths with very low storage requirements. The choice of the "correction" code depends only on the selected delay, making it easy to use. Since multiple delay options are available, conditions are created for merging multiple frequency offset estimates with different delays. By synthesizing several different predetermined delays, a more robust frequency offset estimate can be obtained. To estimate carrier frequency offset (CFO) more accurately and robustly, a multi-delay result merging scheme is adopted.

[0121] Furthermore, the frequency offset estimation range supported by this invention exceeds ±100kHz, and can even be larger; the delay can be set as needed to 1 / 11us, 2 / 11us, ..., 1us. Under the condition of 11MHz sampling rate, that is, corresponding to delays of 1, 2, ..., 11 sampling points, the theoretically supported frequency offset estimation ranges are ±5.5MHz, ±2.75MHz, ..., ±500kHz.

[0122] In addition, this invention requires no preprocessing, does not require oversampling of the original sequence, and only uses the conjugate multiplication of the original sequence with different delay sequences to obtain the sequence. After correction, the sequences are accumulated and calculated to obtain a reliable measure of the frequency offset under that delay value. By increasing the number of collected signal sampling points, the frequency offset estimation performance can be continuously improved. The number of signal sampling points can be any number (N) bits for a duration, and the computational complexity increases linearly with N. Since the overall processing complexity of this invention is not high, and it can achieve near real-time processing to a certain extent, it can accurately estimate the carrier frequency offset within the time observation window during which the signals are roughly synchronized.

[0123] Figure 9 The simulation results provided by this invention are shown in the schematic diagram. Figure 9 As shown, under additive white Gaussian noise conditions, frequency offset estimation is performed using four consecutive preamble bits within a 15μs time frame for different signal-to-noise ratio (SNR) settings. Each sliding step is 1 bit in duration (1μs). The frequency offset estimation method for large frequency offset scenarios provided in this invention is employed, merging the estimation results for delays of 8 and 6. Simulation results show that although frequency offset estimation is performed without complete synchronization, the algorithm incorporates simultaneous coarse synchronization detection and frequency offset estimation, utilizes the unique patterns of different delays, and weakens the influence of unknown preamble bits on the reversal of the frequency offset phase through simplified comparison search. Furthermore, merging the frequency offset phase estimates based on different delays further reduces the impact of noise on frequency offset estimation.

[0124] The frequency offset estimation device for large frequency offset scenarios provided by the present invention is described below. The frequency offset estimation device for large frequency offset scenarios described below can be referred to in correspondence with the frequency offset estimation method for large frequency offset scenarios described above.

[0125] Figure 10 This is a schematic diagram of the frequency offset estimation device for large frequency offset scenarios provided by the present invention, as shown below. Figure 10As shown, this invention provides a frequency offset estimation device for scenarios with large frequency offset, including a detection window data selection module 1001, an accumulation and summation module 1002, an extreme value selection module 1003, and a frequency offset phase angle estimation module 1004. The detection window data selection module 1001 is used to perform windowing processing on the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​based on a preset detection time window, obtaining multiple window data corresponding to the delay conjugate multiplication sequence; the accumulation and summation module 1002 is used to perform accumulation and summation processing on each of the window data, obtaining the accumulation and summation result corresponding to each of the window data; the extreme value selection module 1003 is used to determine the optimal frequency offset phase metric corresponding to different target delay values ​​based on the accumulation and summation result; the frequency offset phase angle estimation module 1004 is used to merge and calculate the optimal frequency offset phase metrics corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence.

[0126] The frequency offset estimation device for large frequency offset scenarios provided by this invention performs delay conjugate multiplication on the signal sequence according to different target delay values ​​through a preset detection time window, and then performs windowing processing to obtain multiple window data; then, the window data are accumulated and summed, and the optimal frequency offset phase metric corresponding to different target delay values ​​is determined based on the results; finally, the optimal frequency offset phase metric of each target delay value is calculated by merging, thereby obtaining the frequency offset estimate of the signal sequence, effectively improving the accuracy of carrier frequency offset estimation.

[0127] The apparatus provided in this embodiment of the invention is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0128] Figure 11 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 11As shown, the electronic device may include: a processor 1101, a communications interface 1102, a memory 1103, and a communication bus 1104, wherein the processor 1101, the communications interface 1102, and the memory 1103 communicate with each other through the communication bus 1104. The processor 1101 can call logical instructions in the memory 1103 to execute a frequency offset estimation method for large frequency offset scenarios. The method includes: performing windowing processing on the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​based on a preset detection time window to obtain multiple window data corresponding to the delay conjugate multiplication sequence; performing cumulative summation processing on each of the window data to obtain the cumulative summation result corresponding to each of the window data; determining the optimal frequency offset phase metric corresponding to different target delay values ​​based on the cumulative summation result; and performing combined calculation on the optimal frequency offset phase metrics corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence.

[0129] Furthermore, the logical instructions in the aforementioned memory 1103 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0130] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, when the program instructions are executed by a computer, the computer is able to execute the frequency offset estimation method for large frequency offset scenarios provided by the above methods, the method including: performing windowing processing on the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​based on a preset detection time window, to obtain multiple window data corresponding to the delay conjugate multiplication sequence; performing cumulative summation processing on each of the window data to obtain the cumulative summation result corresponding to each of the window data; determining the optimal frequency offset phase metric corresponding to different target delay values ​​based on the cumulative summation result; and performing combined calculation on the optimal frequency offset phase metrics corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence.

[0131] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the frequency offset estimation method for large frequency offset scenarios provided in the above embodiments. The method includes: performing windowing processing on the delay conjugate multiplication sequence of a signal sequence under different target delay values ​​based on a preset detection time window to obtain multiple window data corresponding to the delay conjugate multiplication sequence; performing cumulative summation processing on each of the window data to obtain a cumulative summation result corresponding to each of the window data; determining the optimal frequency offset phase metric corresponding to different target delay values ​​based on the cumulative summation result; and performing a combined calculation on the optimal frequency offset phase metrics corresponding to each of the target delay values ​​to obtain a frequency offset estimate of the signal sequence.

[0132] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0133] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0134] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A frequency offset estimation method for scenarios with large frequency offset, characterized in that, include: Based on a preset detection time window, the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​is windowed to obtain multiple windowed data corresponding to the delay conjugate multiplication sequence. The data from each window are summed to obtain the summation result corresponding to each window data. Based on the summation result, the optimal frequency offset phase metric corresponding to different target delay values ​​is determined; The optimal frequency offset phase metric corresponding to each of the target delay values ​​is combined and calculated to obtain the frequency offset estimate of the signal sequence.

2. The frequency offset estimation method for large frequency offset scenarios according to claim 1, characterized in that, The method further includes: Based on a preset set of delay values, a delay conjugate multiplication operation is performed on the signal sequence to obtain the delay conjugate multiplication sequence corresponding to different target delay values; Each delay value in the preset delay value set corresponds to a correction code.

3. The frequency offset estimation method for large frequency offset scenarios according to claim 2, characterized in that, The step of summing up the data from each window to obtain the summation result corresponding to each window data includes: Obtain the target correction code and the corresponding index set; wherein, the index set represents the set of positions of each group to be calculated in the correction code; Based on the target correction code and the index set, the data of each window is accumulated and summed to obtain the accumulated summation result corresponding to each window data.

4. The frequency offset estimation method for large frequency offset scenarios according to claim 1, characterized in that, The step of determining the optimal frequency offset phase metric corresponding to different target delay values ​​based on the accumulated summation result includes: Based on the summation result of the largest modulus among all the windowed data of each of the delay conjugate multiplication sequences, the optimal frequency offset phase metric corresponding to different target delay values ​​is determined.

5. The frequency offset estimation method for large frequency offset scenarios according to claim 1 or 3, characterized in that, The method further includes: The summation result is cached based on a preset cache duration, wherein the preset cache duration is determined based on the length of the preset detection time window.

6. The frequency offset estimation method for large frequency offset scenarios according to claim 1, characterized in that, The step of merging and calculating the optimal frequency offset phase metric corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence includes: The weighted phase term is calculated based on the magnitude, phase angle and target delay value corresponding to each of the optimal frequency offset phase metrics; The weighted phase terms corresponding to all the target delay values ​​are summed to obtain the merged phase metric values. Based on the combined phase metric values ​​and the baseband sampling frequency, the frequency offset estimate of the signal sequence is calculated.

7. A frequency offset estimation device for scenarios with large frequency offset, characterized in that, include: The detection window data selection module is used to perform windowing processing on the delay conjugate multiplication sequence of the signal sequence under different target delay values ​​based on a preset detection time window, so as to obtain multiple window data corresponding to the delay conjugate multiplication sequence. The cumulative summation module is used to perform cumulative summation processing on the data of each window to obtain the cumulative summation result corresponding to each window data. The extreme value selection module is used to determine the optimal frequency offset phase metric corresponding to different target delay values ​​based on the summation result. The frequency offset phase angle estimation module is used to merge and calculate the optimal frequency offset phase metric corresponding to each of the target delay values ​​to obtain the frequency offset estimate of the signal sequence.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the frequency offset estimation method for large frequency offset scenarios as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the frequency offset estimation method for large frequency offset scenarios as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the frequency offset estimation method for large frequency offset scenarios as described in any one of claims 1 to 6.

Images