A constant correlation and low sample number based doppler signal fast detection method

Abstract

The application discloses a kind of based on constant correlation and low sample number Doppler signal fast detection method, comprising: the signal of receiving end collection is carried out Hilbert transform, obtains analytic form received signal, and with the same down-sampling multiple of the constant correlation detection template constructed in equal interval extraction, obtain low sample detection signal sequence;With the low sample template signal in constant correlation detection template, the normalized cross-correlation calculation of down-sampled received signal is carried out, extracts the maximum peak value of correlation function and its corresponding Doppler compression ratio;According to the Doppler compression ratio corresponding to maximum correlation peak, the Doppler factor and Doppler frequency offset parameter of signal are calculated, and the detection result is output, realize the fast estimation of underwater acoustic signal Doppler signal.The down-sampling process of the application will not change the amplitude and position of cross-correlation main peak, so as to significantly reduce the sample number and the calculation complexity, while maintaining the detection precision and Doppler signal detection performance.

Description

Technical Field

[0001] This invention belongs to the field of underwater acoustic communication technology, and relates to Doppler signal detection technology, specifically to a rapid Doppler signal detection method based on constant correlation and low sample number. Background Technology

[0002] Doppler signal detection is a crucial technology in underwater acoustic communication and positioning systems. Due to the relative motion between the transmitter and receiver, the propagating signal undergoes time-scale changes and frequency shifts, leading to increased demodulation error rates and synchronization failures. Accurate estimation of the Doppler factor is essential for recovering carrier phase, maintaining symbol synchronization, and improving system reliability.

[0003] Currently, commonly used Doppler detection methods mainly include the Ambiguity Function Method (AFM) and the Block Doppler Compensation (BDC). The AFM method, through a two-dimensional search in terms of time delay and Doppler dimensions, can comprehensively reflect the time-frequency characteristics of the signal, but its computational load is extremely large, making it difficult to meet real-time processing requirements. The BDC method processes the received signal in segments and calculates the cross-correlation of the blocks to estimate local Doppler changes. Although this reduces the computational load to some extent, it still suffers from computational complexity and long processing time in scenarios with high sampling rates and long synchronization header signals.

[0004] Therefore, a new technological solution is needed to address these issues. Summary of the Invention

[0005] Objective: To overcome the shortcomings of existing Doppler detection methods, such as dependence on high sampling rates, high computational complexity, and insufficient real-time performance, this invention provides a rapid Doppler signal detection method based on constant correlation and low sample number. The downsampling process does not alter the amplitude and position of the main cross-correlation peak, thus significantly reducing the number of sample points and computational complexity while maintaining detection accuracy and Doppler signal detection performance. This invention features simple structure, strong real-time performance, and requires no additional hardware support, making it suitable for applications with limited storage and computing resources, such as underwater communication, acoustic ranging, and target localization.

[0006] Technical Solution: To achieve the above objectives, this invention provides a rapid detection method for Doppler signals based on constant correlation and low sample number, comprising the following steps:

[0007] S1: Perform Hilbert transform on the signal acquired by the receiving end to obtain the analytical form of the received signal, and extract it at equal intervals with the same downsampling factor in the constructed constant correlation detection template to obtain the low sample point detection signal sequence.

[0008] S2: Perform normalized cross-correlation calculation between the downsampled received signal and the low-sample template signal in the constant correlation detection template, and extract the maximum peak value of the correlation function and its corresponding Doppler compression ratio;

[0009] S3: Based on the Doppler compression ratio corresponding to the maximum correlation peak, calculate the Doppler factor and Doppler frequency offset parameters of the signal, and output the detection results to achieve rapid estimation of the underwater acoustic signal Doppler signal.

[0010] Furthermore, the construction of the constant correlation detection template in step S1 includes:

[0011] A1: Perform Hilbert transform on the original synchronization head signal to obtain the analytical form of the master template signal, and unify its time axis to provide a reference signal for the subsequent construction of the low-sample template library;

[0012] A2: Based on the preset Doppler compression ratio, the master template signal is stretched or compressed on a time scale, and the transformed signal is extracted at equal intervals with a fixed downsampling factor to generate low-sample template signals, which constitute the constant correlation detection template library.

[0013] Furthermore, in step A1, the Hilbert transform is used to convert the real signal into an analytic signal in order to extract the envelope and phase information of the signal and improve the accuracy of subsequent correlation detection.

[0014] Furthermore, in step A1, the time axis of the master template signal is standardized according to the system sampling rate to ensure the consistency of the template signal and the received signal on the time reference and avoid correlation peak shift caused by sampling differences.

[0015] Furthermore, in step A2, the Doppler compression ratio is determined by an interpolation factor. It means, set ;in, The relative velocity between the transmitter and receiver. The velocity of sound is constant; the Doppler compression ratio is... Used to describe the degree of stretching and compression of a signal in the time domain, when Time stretching of the signal corresponds to Doppler compression; when The signal is compressed in time, corresponding to Doppler expansion.

[0016] Furthermore, in step A2, the time scale stretching or compression is achieved through interpolation; the interpolated time series is represented as follows: ,in, This is the time sequence of the original synchronization header signal. The time series is after Doppler scaling transformation; the synchronization head signal is on the new time axis. Interpolation calculations are performed to obtain the Doppler-corrected template signal. The interpolation method can be linear interpolation or spline interpolation to ensure the phase continuity and amplitude smoothness of the signal, thereby forming a low-sample template library that can maintain constant correlation under different downsampling ratios.

[0017] Furthermore, in step A2, the downsampling operation employs a fixed-step, equally spaced decimation method to maintain the phase continuity and amplitude ratio of the signal. Under the premise of satisfying the bandpass sampling condition, no anti-aliasing filter is required, and the downsampling factor is [value missing]. The generated template signals are stored in an offline template library for subsequent constant correlation matching detection. This downsampling strategy ensures that the amplitude and position of the normalized cross-correlation main peak remain consistent at low sampling rates, thereby maintaining detection accuracy and consistency with Doppler estimation while significantly reducing the number of samples and computational load.

[0018] Furthermore, the normalized cross-correlation calculation in step S2 is used to measure the similarity between the received signal and the template signal, and its calculation form is expressed as follows: ;in, The template signal after downsampling. The received signal after downsampling. This is a delayed variable. Through this normalization process, the influence of different sampling rates or signal amplitude differences on the correlation results can be eliminated, thereby achieving consistency in the amplitude and position of the main correlation peak under different downsampling rates. This enables the constant correlation detection characteristic, achieving the constant correlation detection effect. The constant correlation detection characteristic ensures that the detection results under low sample point conditions can reproduce the matching performance under the original high sampling rate.

[0019] Furthermore, in step S2, the main peak detection is performed by searching a correlation function. The maximum value is achieved, and the corresponding delay position and Doppler compression ratio are extracted as parameters for the best matching between the received signal and the template signal, which are used for the rapid identification and estimation of the Doppler effect.

[0020] Furthermore, in step S3, the Doppler factor is determined based on the normalized cross-correlation function. The Doppler compression ratio corresponding to the maximum peak value Calculations show that, let's assume ;in, The Doppler factor of the signal represents the compression or expansion ratio of the received signal to the original template signal over the time scale.

[0021] Furthermore, the Doppler frequency offset parameter in step S3 is calculated from the Doppler factor and is expressed as follows: ;in, The system center frequency, This represents the corresponding frequency offset. By extracting the position of the main peak of the normalized cross-correlation and its corresponding Doppler compression ratio, the Doppler factor and frequency offset parameters of the signal can be quickly obtained, enabling real-time estimation and detection output of the Doppler effect of underwater acoustic signals.

[0022] Beneficial effects: Compared with existing technologies, this invention can still extract accurate parameters for calculation while significantly reducing the amount of data, thus achieving real-time estimation. This invention has the following advantages:

[0023] 1. By introducing the Hilbert transform to obtain the analytical form of the synchronization head signal, the combined use of envelope and phase is realized, which significantly improves the accuracy of signal matching detection;

[0024] 2. An interpolation method based on Doppler compression ratio is adopted to stretch or compress the synchronization head signal on a time scale, and a low-sample template library is constructed by combining an equal-interval downsampling strategy, which greatly reduces the amount of computation while maintaining the consistency of the amplitude of the relevant main peaks;

[0025] 3. By using normalized cross-correlation calculation to achieve constant correlation detection, the amplitude and position of the correlation peak are kept consistent under different downsampling ratios, thereby effectively avoiding detection deviation caused by energy normalization error;

[0026] 4. Based on the extraction of Doppler factor and frequency offset parameters from the maximum correlation peak, the Doppler effect of the signal can be estimated quickly and accurately, enabling real-time detection and tracking of underwater acoustic signals;

[0027] 5. The overall structure is simple and easy to implement. It can significantly improve the detection speed and real-time performance of the system without increasing hardware complexity, and has high engineering application value. Attached Figure Description

[0028] Figure 1 This is a schematic flowchart of the method of the present invention;

[0029] Figure 2 Plots showing the normalized cross-correlation results under different downsampling factors;

[0030] Figure 3 A comparison of the normalized spectrum of the signal under different downsampling factors;

[0031] Figure 4 The graph shows the change in detection probability as a function of signal-to-noise ratio under different downsampling factors;

[0032] Figure 5 This is a comparison chart of the average runtime under different downsampling factors. Detailed Implementation

[0033] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading this invention, any modifications of the invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0034] Example 1:

[0035] like Figure 1 As shown, this embodiment provides a fast Doppler signal detection method based on constant correlation and low sample number, including two main stages: offline template library generation and online detection. Specifically, it includes the following steps:

[0036] I. Offline Template Library Generation

[0037] The construction of the constant correlation detection template includes:

[0038] A1: Perform Hilbert transform on the original synchronization head signal to obtain the analytical form of the master template signal, and unify its time axis to provide a reference signal for the subsequent construction of the low-sample template library;

[0039] The Hilbert transform is used to convert real signals into analytic signals in order to extract the envelope and phase information of the signal and improve the accuracy of subsequent correlation detection.

[0040] The time axis of the master template signal is standardized according to the system sampling rate to ensure the consistency of the template signal and the received signal in terms of time reference and to avoid correlation peak shift caused by sampling differences.

[0041] A2: Based on the preset Doppler compression ratio, the master template signal is stretched or compressed on a time scale, and the transformed signal is extracted at equal intervals with a fixed downsampling factor to generate low-sample template signals, which constitute the constant correlation detection template library.

[0042] Doppler compression ratio via interpolation factor It means, set ;in, The relative velocity between the transmitter and receiver. The velocity of sound is constant; the Doppler compression ratio is... Used to describe the degree of stretching and compression of a signal in the time domain, when Time stretching of the signal corresponds to Doppler compression; when The signal is compressed in time, corresponding to Doppler expansion.

[0043] Time-scale stretching or compression is achieved through interpolation; the interpolated time series is represented as follows: ,in, This is the time sequence of the original synchronization header signal. The time series is after Doppler scaling transformation; the synchronization head signal is on the new time axis. Interpolation calculations are performed to obtain the Doppler-corrected template signal. The interpolation method can be linear interpolation or spline interpolation to ensure the phase continuity and amplitude smoothness of the signal, thereby forming a low-sample template library that can maintain constant correlation under different downsampling ratios.

[0044] The downsampling operation employs a fixed-step, equally spaced decimation method to maintain the phase continuity and amplitude ratio of the signal. Provided the bandpass sampling conditions are met, no anti-aliasing filter is required, and the downsampling factor is [value missing]. The generated template signals are stored in an offline template library for subsequent constant correlation matching detection. This downsampling strategy ensures that the amplitude and position of the normalized cross-correlation main peak remain consistent at low sampling rates, thereby maintaining detection accuracy and consistency with Doppler estimation while significantly reducing the number of samples and computational load.

[0045] II. Online Testing

[0046] 1) Perform Hilbert transform on the signal collected by the receiving end to obtain the analytical form of the received signal, and extract it at equal intervals with the same downsampling factor in the constructed constant correlation detection template to obtain the low sample point detection signal sequence;

[0047] 2) Perform normalized cross-correlation calculation between the downsampled received signal and the low-sample template signal in the constant correlation detection template, and extract the maximum peak value of the correlation function and its corresponding Doppler compression ratio;

[0048] Normalized cross-correlation is used to measure the similarity between the received signal and the template signal. Its calculation form is expressed as follows: ;in, The template signal after downsampling. The received signal after downsampling. This is a time delay variable, representing the amount of time displacement of the received signal.

[0049] This normalization process eliminates the influence of different sampling rates or signal amplitude differences on the correlation results, thereby achieving consistency in the amplitude and position of the main correlation peak under different downsampling rates. This enables the realization of constant correlation detection characteristics, achieving the desired constant correlation detection effect. The constant correlation detection characteristic ensures that the detection results under low sample point conditions can reproduce the matching performance under the original high sampling rate.

[0050] Main peak detection is achieved by searching for relevant functions. The maximum value is achieved, and the corresponding delay position and Doppler compression ratio are extracted as parameters for the best matching between the received signal and the template signal, which are used for the rapid identification and estimation of the Doppler effect.

[0051] 3) Based on the Doppler compression ratio corresponding to the maximum correlation peak, calculate the Doppler factor and Doppler frequency offset parameters of the signal, and output the detection results to achieve rapid estimation of the underwater acoustic signal Doppler signal.

[0052] Doppler factor based on normalized cross-correlation function The Doppler compression ratio corresponding to the maximum peak value Calculations show that, let's assume ;in, The Doppler factor of the signal represents the compression or expansion ratio of the received signal to the original template signal over the time scale.

[0053] The Doppler frequency offset parameter is calculated from the Doppler factor and is expressed as: ;in, The system center frequency, This represents the corresponding frequency offset.

[0054] By extracting the position of the main peak of the normalized cross-correlation and its corresponding Doppler compression ratio, the Doppler factor and frequency offset parameters of the signal can be obtained quickly, enabling real-time estimation and output of the Doppler effect of underwater acoustic signals.

[0055] Example 2:

[0056] This embodiment applies the method provided in Embodiment 1 in a specific way, and the process includes:

[0057] I. Offline Template Library Generation

[0058] In this embodiment, the system sampling rate is set to To verify the performance of the constant correlation detection method under different downsampling conditions, a downsampling factor was selected. These correspond to the original sampling rate and the multiple downsampling scenarios, respectively. The Doppler compression ratio is set to... ,in The range of values ​​is Step size is The synchronization header signal is processed by Hilbert transform to obtain its analytical form. This signal contains both envelope and phase information, providing a precise basis for subsequent interpolation and cross-correlation operations.

[0059] During the template library construction phase, the system parses signals. According to different Doppler ratios Perform a time-scale transformation to obtain the stretched or compressed signal: Subsequently, the transformed signal undergoes interpolation on the time axis, using either spline interpolation or linear interpolation to maintain signal continuity and smoothness. After interpolation, the signal is decimated at fixed intervals to form a downsampled low-sample template. .in This is the downsampling factor. This downsampling process does not require an additional anti-aliasing filter. As long as the bandpass sampling condition is met, the amplitude and position of the normalized cross-correlation main peak can be maintained while significantly reducing the number of samples.

[0060] Repeating the above operations for all candidate Doppler ratios and different downsampling rates yields a constant correlation detection template library covering the target search range. This template library is directly loaded and used in subsequent detection stages, thus avoiding repeated generation processes, significantly reducing computational burden and improving detection real-time performance.

[0061] II. Online Testing

[0062] During the online detection phase, the system first loads the pre-built constant correlation template library and its parameter set. The received signal consists of a known synchronization header and a data segment, with the synchronization header used for Doppler estimation. The signal center frequency is set in the simulation scenario. The signal sampling rate was 96kHz. To evaluate the robustness of the algorithm under different noise conditions, the signal-to-noise ratio was set from -25dB to 15dB, and multiple Monte Carlo experiments were conducted independently at each signal-to-noise ratio.

[0063] The transmitted signal is composed of a synchronization header and a data segment, and its analytical form is obtained through Hilbert transform for complex domain correlation calculations. The signal propagates through a simulated underwater acoustic channel, the channel model of which includes path delay, amplitude attenuation, and Doppler compression factor. To simulate the Doppler effect caused by motion, a time-scale transformation is performed on the transmitted signal. The signal is then superimposed onto the receiving end time series according to the set time delay to form a received signal containing Doppler stretching and propagation delay.

[0064] In noisy environments, complex Gaussian white noise is added to the received signal, with its power calculated based on the target signal-to-noise ratio. After obtaining the noisy signal, a time segment including the synchronization header is extracted and used as the detection input. To ensure consistency with the template library, the detection signal is downsampled at the current downsampling rate. Equal-interval sampling is performed to form a low-sample detection sequence.

[0065] The system performs normalized cross-correlation calculations between the detected sequence and each candidate template at the corresponding fold in the template library:

[0066]

[0067] And record the Doppler ratios. The corresponding maximum correlation amplitude And calculate the estimated Doppler factor. Ultimately, this enables rapid detection of the Doppler factor.

[0068] The Doppler frequency offset parameter is calculated from the Doppler factor and is expressed as: ;in, The system center frequency, This represents the corresponding frequency offset.

[0069] By extracting the position of the main peak of the normalized cross-correlation and its corresponding Doppler compression ratio, the Doppler factor and frequency offset parameters of the signal can be obtained quickly, enabling real-time estimation and output of the Doppler effect of underwater acoustic signals.

[0070] Example 3:

[0071] To verify the effectiveness of the method of the present invention, the following performance comparison data were obtained through simulation experiments in this embodiment:

[0072] like Figure 2 The figure shows the time-domain cross-correlation results of the received signals under different downsampling factors. Figure 2 The normalized cross-correlation functions after downsampling at the original sampling rates of 96 kHz and 48 kHz, 24 kHz, and 12 kHz are presented respectively. It can be seen that under each sampling rate condition, the amplitude and position of the main peak of the correlation function remain consistent, and the sidelobe distribution changes little. This indicates that the equal-interval downsampling of the present invention does not destroy the correlation main peak criterion, thus maintaining detection accuracy and Doppler estimation stability while reducing the number of sampling points.

[0073] like Figure 3 The corresponding frequency domain characteristic diagram is shown below. Figure 3 The normalized spectral distribution of the signal at different sampling rates is presented. It can be seen that the main lobe morphology and energy distribution after downsampling are basically consistent with the original sampled signal, without significant spectral broadening or distortion. This result further verifies that the constant correlation detection characteristic of the method of this invention also holds in the frequency domain when the bandpass sampling condition is met, providing a reliable guarantee for the rapid detection and robust estimation of Doppler signals.

[0074] like Figure 4 The figure shows the detection probability curves of the method of the present invention at different downsampling factors. The horizontal axis represents the signal-to-noise ratio (SNR, in dB), and the vertical axis represents the detection rate. Figure 4It is evident that under low signal-to-noise ratio (SNR below -20 dB), a higher downsampling rate results in a slight decrease in detection rate; however, when the SNR exceeds -15 dB, the curves at each downsampling rate gradually converge, achieving a 100% detection probability. This result demonstrates that the constant correlation low-sample-point detection method proposed in this invention maintains detection performance comparable to the full-sample cross-correlation method under different downsampling conditions, while significantly reducing computational complexity and the number of sample points, thus achieving a balance between detection efficiency and accuracy.

[0075] like Figure 5 The figure shows the comparison of the average running time of the method of the present invention under different downsampling factors. The horizontal axis represents the detection method category, and the vertical axis represents the average running time (unit: ms). Figure 5 As can be seen, compared with traditional cross-correlation methods, the constant correlation low-sample-point detection method of this invention reduces the average running time by approximately 50%, 75%, and 85% under downsampling conditions of 2x, 4x, and 8x, respectively, and the computational efficiency significantly improves with increasing downsampling factor. These results demonstrate that this invention effectively reduces computational complexity and processing time while maintaining stable detection performance, significantly improving the real-time performance and engineering applicability of Doppler detection.

Claims

1. A rapid detection method for Doppler signals based on constant correlation and low sample number, characterized in that, Includes the following steps: S1: Perform Hilbert transform on the signal acquired by the receiving end to obtain the analytical form of the received signal, and extract it at equal intervals with the same downsampling factor in the constructed constant correlation detection template to obtain the low sample point detection signal sequence. S2: Perform normalized cross-correlation calculation between the downsampled received signal and the low-sample template signal in the constant correlation detection template, and extract the maximum peak value of the correlation function and its corresponding Doppler compression ratio; S3: Based on the Doppler compression ratio corresponding to the maximum correlation peak, calculate the Doppler factor and Doppler frequency offset parameters of the signal, and output the detection results to achieve rapid estimation of the underwater acoustic signal Doppler signal.

2. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 1, characterized in that, The construction of the constant correlation detection template in step S1 includes: A1: Perform Hilbert transform on the original synchronization head signal to obtain the analytical form of the master template signal, and unify its time axis to provide a reference signal for the subsequent construction of the low-sample template library; A2: Based on the preset Doppler compression ratio, the master template signal is stretched or compressed on a time scale, and the transformed signal is extracted at equal intervals with a fixed downsampling factor to generate low-sample template signals, which constitute the constant correlation detection template library.

3. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 2, characterized in that, In step A2, the Doppler compression ratio is determined by an interpolation factor. It means, set ;in, The relative velocity between the transmitter and receiver. The velocity of sound is constant; the Doppler compression ratio is... Used to describe the degree of stretching and compression of a signal in the time domain, when Time stretching of the signal corresponds to Doppler compression; when The signal is compressed in time, corresponding to Doppler expansion.

4. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 3, characterized in that, In step A2, time-scale stretching or compression is achieved through interpolation; the interpolated time series is represented as follows: ,in, This is the time sequence of the original synchronization header signal. The time series is after Doppler scaling transformation; the synchronization head signal is on the new time axis. Interpolation calculations are performed to obtain the Doppler-corrected template signal. .

5. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 2, characterized in that, In step A2, the downsampling operation employs a fixed-step, equally spaced decimation method to maintain the phase continuity and amplitude ratio of the signal. The downsampling factor is [missing value]. The generated template signals are stored in an offline template library for subsequent constant correlation matching detection.

6. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 1, characterized in that, The normalized cross-correlation calculation in step S2 is used to measure the similarity between the received signal and the template signal, and its calculation form is expressed as follows: ;in, The template signal after downsampling. The received signal after downsampling. It is a delayed variable.

7. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 6, characterized in that, In step S2, the main peak detection is performed by searching a correlation function. The maximum value is achieved, and the corresponding delay position and Doppler compression ratio are extracted as parameters for the best matching between the received signal and the template signal, which are used for the rapid identification and estimation of the Doppler effect.

8. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 7, characterized in that, In step S3, the Doppler factor is determined based on the normalized cross-correlation function. The Doppler compression ratio corresponding to the maximum peak value Calculations show that, let's assume ;in, The Doppler factor of the signal represents the compression or expansion ratio of the received signal to the original template signal over the time scale.

9. The method for rapid detection of Doppler signals based on constant correlation and low sample number as described in claim 8, characterized in that, In step S3, the Doppler frequency offset parameter is calculated from the Doppler factor and expressed as follows: ;in, The system center frequency, This represents the corresponding frequency offset.

Images