A method and system for receiving and processing VLF / LF signals

Abstract

The present application relates to the technical field of VLF / LF communication, and discloses a VLF / LF signal receiving and processing method and system, which adopts multiple miniaturized VLF / LF antennas and front ends to form a receiving array for jointly receiving VLF / LF band signals, and based on a multi-channel signal iterative correlation structure, integrally implements functions such as blind signal synthesis and signal detection, and can obtain good signal distribution detection performance in the whole VLF / LF receiving band and high-quality signal waveform samples under the condition that the signal spectrum distribution information in the whole VLF / LF receiving band is unknown and different array layouts are adopted. The method and system fully transfer and utilize the information and signals extracted by the two processing links of signal detection and signal synthesis, and therefore can obtain better processing performance.

Description

Technical Field

[0001] This invention belongs to the field of VLF / LF communication technology, and particularly relates to a VLF / LF signal receiving and processing method and system. Background Technology

[0002] Time-frequency segmentation is a fundamental step in communication signal reception and processing, and the rationality of segmentation directly affects processing performance. Frequency domain segmentation is generally based on the distribution information of the frequency domain positions of different signals within the received frequency band; time domain segmentation mainly considers the trade-off between processing performance and complexity in stationary noise environments, while in non-stationary noise environments, non-stationary characteristics need to be considered.

[0003] The VLF / LF band has a low frequency range. When using miniaturized antennas and front-end receivers, the noise in the received signal comes not only from receiver thermal noise but also from ambient noise induced by the antenna, resulting in non-stationary noise characteristics. When multiple antenna front-ends are used to jointly receive signals, the VLF / LF band noise received by different antenna front-ends exhibits correlation within a certain spatial, temporal, and frequency range. All of these factors affect the VLF / LF signal reception and processing performance.

[0004] Processes similar to this invention mainly involve signal detection and signal synthesis.

[0005] Typical signal detection methods mainly include energy-based detection methods and signal feature-based detection methods. The basic principle and steps of energy-based signal detection methods can be described as follows:

[0006] Step ①: Calculate the energy or average power of a signal sample in a specific time / frequency band as a detection statistic;

[0007] Step 2: Set the detection threshold;

[0008] Step 3: Compare the detection statistic obtained in Step 1 with the detection threshold obtained in Step 2. If it is greater than the threshold, the detection result is "signal present"; if it is less than the threshold, the detection result is "no signal present".

[0009] A typical example of a signal feature-based detection method is the matched filtering-based signal detection method, whose basic principle and steps can be described as follows:

[0010] Step ①: Design a matched filter based on the known signal waveform function;

[0011] Step 2: Pass the signal samples of a specific time and frequency band through the matched filter designed in Step 1, and select the maximum value of the matched filter output as the detection statistic;

[0012] Step 3: Set the detection threshold;

[0013] Step 4: Compare the maximum value of the matched filter output obtained in step 2 with the detection threshold obtained in step 3. If it is greater than the threshold, the detection result is "signal present"; if it is less than the threshold, the detection result is "no signal present".

[0014] Signal synthesis refers to the use of multiple antennas to jointly receive the same signal or signals within the same frequency band. Typical signal synthesis methods mainly include time-domain synthesis and frequency-domain synthesis. The basic principles and steps of time-domain synthesis can be described as follows:

[0015] Step ①: Select one signal from the multiple signals received from multiple antennas or generate one signal based on a specific method as a reference signal;

[0016] Step 2: Based on signal samples of a given duration, estimate the differences in parameters such as propagation delay, carrier phase, and carrier frequency between each signal and the reference signal obtained in Step 1;

[0017] Step ③: Using the reference signal from Step ① as a benchmark, and based on the differences in the parameters obtained in Step ②, compensate for the time delay, phase, and frequency difference of each signal received by multiple antennas;

[0018] Step 4: After compensating for the differences in parameters in Step 3, add the signals of each channel with specific weights to obtain the composite signal.

[0019] The basic principles and steps of the frequency domain synthesis method can be described as follows:

[0020] Step ①: Based on the given duration and frequency band parameters, within the processing frequency band, transform the multiple signals obtained from multiple antennas by sampling from the time domain to the frequency domain using Fourier transform;

[0021] Step 2: Select one signal from the multiple signals received from multiple antennas or generate one signal based on a specific method as a reference signal;

[0022] Step 3: Based on the signal samples of the given frequency band and duration obtained in Step 2, estimate the differences in parameters such as phase and frequency between each signal and the corresponding frequency signal component of the reference signal obtained in Step 2;

[0023] Step 4: Using the reference signal from Step 2 as a benchmark, and based on the parameter differences of the signal components at each frequency point obtained in Step 3, compensate for the phase and frequency differences of the corresponding signal components at each frequency point in the frequency domain for each signal sampled by multiple antennas.

[0024] Step 5: After compensating for the parameter differences in Step 4, the signals of each channel are weighted with specific weights and then added together to obtain the frequency domain composite signal. The signal is then transformed into the time domain composite signal through inverse Fourier transform.

[0025] In addition, similar methods include signal detection based on multi-antenna joint reception, with typical methods as follows. The basic idea of ​​typical method 1 is to perform signal detection processing on each signal received by multiple antennas and output the detection result indicating the presence or absence of a signal. A large-number decision is applied to the multiple detection results, i.e., the decision rule is: if more than N / 2 of the N detection results indicate "signal presence," then the signal is considered "present"; otherwise, it is considered "no signal." The implementation block diagram of the above method is shown below. Figure 1 As shown.

[0026] The basic idea of ​​typical method 2 is to first synthesize the signals received by multiple antennas to obtain a synthesized signal sample with a higher signal-to-noise ratio, and then perform signal detection based on the synthesized signal, outputting the detection result. The implementation block diagram of this method is shown below. Figure 2 As shown.

[0027] The main drawbacks of the existing methods and systems described above are as follows:

[0028] ① It is only applicable to the detection and processing of single signals, and requires a certain amount of prior information to achieve good processing performance.

[0029] For example, energy-based signal detection methods require prior knowledge of the frequency band of the signal to determine its presence or absence. If the detection bandwidth contains multiple signals, it cannot determine the presence of a specific signal. Matched filtering-based signal detection requires prior knowledge of the waveform of the signal to effectively detect a single signal. If multiple signals exist within the receiving processing frequency band, the above methods lack the general processing capability to detect multiple signals within the same frequency band.

[0030] ②Suitable for stable noise environments; performance degrades under non-stationary noise conditions.

[0031] The methods described above were all obtained under typical stationary noise conditions, such as additive white Gaussian noise. Their performance will inevitably degrade in non-stationary noise environments. Furthermore, when miniaturized antennas and front-ends are used to receive VLF / LF signals, the noise exhibits non-stationary characteristics.

[0032] ③ Signal detection based on multi-antenna joint reception does not fully utilize the signals and information from each antenna.

[0033] For example, in joint signal detection based on large number decision, if the signal-to-noise ratio (SNR) of each individual signal is low, the probability of correct detection for each individual signal will be low, resulting in limited improvement in the correct probability of joint detection. Although joint signal detection based on signal synthesis improves the SNR through signal synthesis, it is only applicable to the detection of a single signal and lacks versatility when dealing with multiple signals within the same frequency band. Summary of the Invention

[0034] To address the shortcomings of existing methods in VLF / LF signal reception and processing applications, the present invention aims to design a collaborative VLF / LF signal reception and processing method and system based on the joint reception of multiple miniaturized antennas and front-ends, employing iterative processing of two stages: VLF / LF full-band signal synthesis enhancement and VLF / LF full-band signal distribution information detection. This method can achieve joint detection of multiple signals within the VLF / LF band, acquisition of frequency domain distribution information, and output of high-quality signal samples. It can achieve better signal distribution detection performance and high-quality signal waveform samples even when the signal spectrum distribution information within the entire VLF / LF receiving band is unknown and different array layouts are used.

[0035] To achieve the above objectives, the present invention adopts the following technical solution:

[0036] This invention provides a VLF / LF signal receiving and processing method, comprising:

[0037] Step 1: Perform multi-channel sampling on the received multiple VLF / LF signals;

[0038] Step 2: Set initial processing parameters, including initial VLF / LF frequency band frequency domain division, initial frequency domain related segment length, and initial time domain related segment length;

[0039] Step 3: Perform integrated processing on the received multi-channel VLF / LF full-band signals;

[0040] Step 4: Detect the signal distribution information of the VLF / LF band;

[0041] Step 5: Iteratively execute steps 3 and 4 until the convergence condition is met. Output the synthesized signal samples of each sub-band containing the signal after inverse Fourier transform, and output the frequency domain distribution information of the received and processed frequency band signal obtained from the detection.

[0042] Further, step 1 includes:

[0043] The received multi-channel VLF / LF signals are subjected to direct sampling transformation across the entire frequency band of [3KHz~300KHz], with a sampling rate of not less than 1MHz.

[0044] Furthermore, the initial VLF / LF band frequency domain division settings include:

[0045] If the frequency domain distribution of the signal within the VLF / LF band is known, the initial VLF / LF band frequency domain division is set according to the signal frequency domain distribution information.

[0046] If the number of signals and frequency domain distribution information within the VLF / LF band are unknown, the initial band division is performed based on the signal reception characteristics of the VLF / LF band.

[0047] If the frequency domain distribution information of the signal within the receiving and processing frequency band is known, then the known noise sub-frequency band flag is set to S = '0', and the initial flags of the remaining sub-frequency bands are all set to S = '#'. S = '0' indicates no signal, and S = '#' indicates no judgment is made for the time being.

[0048] Furthermore, the setting of the initial frequency domain correlation segment length includes:

[0049] If the frequency domain distribution of the signal within the VLF / LF band is known, then Where F i (0) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the initial stage. This represents the frequency division region of the i-th sub-band in the initial stage;

[0050] If the number of signals and frequency domain distribution information within the VLF / LF bands to be received and processed are unknown, then set F i (0) =1.

[0051] Furthermore, the setting of the initial time-domain correlation segment length includes:

[0052] If there is a known frequency region without effective signal in the initial VLF / LF band frequency domain division setting, then the noise samples collected by multiple receiving channels in the frequency region are extracted, the variance characteristics of the noise samples are statistically analyzed, and the initial time domain correlation segment length of each sub-band is set according to the time interval when the noise variance characteristics show significant differences.

[0053] If the frequency regions without effective signals in each sub-band of the initial VLF / LF frequency band frequency domain division setting are unknown, the initial time domain related segment length is set based on empirical values.

[0054] Further, step 3 includes:

[0055] Step 31: Calculate the short-time Fourier transform of each VLF / LF signal in the k-th iteration;

[0056] Step 32: Perform frequency domain correlation processing on each VLF / LF signal and the synthesized signal:

[0057]

[0058]

[0059] in This indicates that the i-th sub-band and j-th path obtained in the k-th iteration is centered at τ and has a duration of T. i k The result of the short-time Fourier transform of the signal, where f is the independent variable of the short-time Fourier transform, and T is the signal. i (k) This represents the time-domain segment length of the i-th sub-band signal in the k-th iteration; This indicates that the i-th sub-band obtained in the (k-1)-th iteration is centered at τ and has a duration of T. i k-1 The frequency domain synthesized signal, T i k-1 This represents the time-domain segment length of the i-th sub-band signal in the (k-1)-th iteration; F represents the frequency division region of the i-th sub-band in the (k-1)-th iteration; i (k-1) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the (k-1)-th iteration; All are process variables; Represents the integral variable;

[0060] Step 33: Perform difference compensation on the frequency domain synthesis parameters of the k-th iteration, and calculate the frequency domain data after parameter difference compensation according to the following formula.

[0061]

[0062] Step 34: Calculate the frequency domain data of the synthesized signal in the k-th iteration.

[0063]

[0064] Step 35: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require the Pass it to step 4.

[0065] Further, step 4 includes:

[0066] Step 41: If the i-th sub-band flag S ≠ '0', then calculate the k-th iteration according to the following formula. Relevant phase spectrum information

[0067]

[0068] in Indicates taking The imaginary part, Indicates taking The real part;

[0069] Step 42: [Regarding...] Perform linear fitting to obtain the fitted linear function.

[0070]

[0071] in It is the slope of the fitted linear function. It is the intercept of the fitted linear function;

[0072] Step 43: Calculation and To determine the error between the two, set a signal detection threshold μ and compare the error with the threshold.

[0073] Step 44: Update the signal frequency domain distribution information. If the error value is less than the threshold, the sub-frequency band division remains unchanged; if the error value is greater than the threshold, the i-th sub-frequency band in the (k-1)-th iteration is evenly divided into two new sub-frequency bands, thus obtaining the sub-frequency band division of the corresponding frequency region in the k-th iteration.

[0074] Step 45: Update the corresponding processing parameters based on the updated sub-band allocation scheme obtained in Step 44, and then update the parameters accordingly. T i (k) F i (k) Pass to step 3, where This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the k-th iteration.

[0075] Another aspect of the present invention provides a VLF / LF signal receiving and processing system, comprising:

[0076] The multi-channel sampling module is used to perform multi-channel sampling on the received multiple VLF / LF signals;

[0077] The initial processing parameter setting module is used to set the initial processing parameters, which include the initial VLF / LF frequency band frequency domain division, the initial frequency domain related segment length, and the initial time domain related segment length.

[0078] The signal synthesis and processing module is used to perform integrated synthesis processing on multiple received VLF / LF full-band signals.

[0079] The signal distribution information detection module is used to detect the signal distribution information of the VLF / LF frequency band;

[0080] The convergence judgment and signal information output module is used to iteratively execute the integrated signal synthesis processing module and the signal distribution information detection module until the convergence condition is met. It outputs the synthesized signal samples of each sub-frequency band of the existing signal after inverse Fourier transform, and outputs the frequency domain distribution information of the received and processed frequency band signal obtained by detection.

[0081] Furthermore, the multi-channel sampling module is specifically used for:

[0082] The received multi-channel VLF / LF signals are subjected to direct sampling transformation across the entire frequency band of [3KHz~300KHz], with a sampling rate of not less than 1MHz.

[0083] Furthermore, the initial VLF / LF band frequency domain division settings include:

[0084] If the frequency domain distribution of the signal within the VLF / LF band is known, the initial VLF / LF band frequency domain division is set according to the signal frequency domain distribution information.

[0085] If the number of signals and frequency domain distribution information within the VLF / LF band are unknown, the initial band division is performed based on the signal reception characteristics of the VLF / LF band.

[0086] If the frequency domain distribution information of the signal within the receiving and processing frequency band is known, then the known noise sub-frequency band flag is set to S = '0', and the initial flags of the remaining sub-frequency bands are all set to S = '#'. S = '0' indicates no signal, and S = '#' indicates no judgment is made for the time being.

[0087] Furthermore, the setting of the initial frequency domain correlation segment length includes:

[0088] If the frequency domain distribution of the signal within the VLF / LF band is known, then Where F i (0) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the initial stage. This represents the frequency division region of the i-th sub-band in the initial stage;

[0089] If the number of signals and frequency domain distribution information within the VLF / LF bands to be received and processed are unknown, then set F i (0) =1.

[0090] Furthermore, the setting of the initial time-domain correlation segment length includes:

[0091] If there is a known frequency region without effective signal in the initial VLF / LF band frequency domain division setting, then the noise samples collected by multiple receiving channels in the frequency region are extracted, the variance characteristics of the noise samples are statistically analyzed, and the initial time domain correlation segment length of each sub-band is set according to the time interval when the noise variance characteristics show significant differences.

[0092] If the frequency regions without effective signals in each sub-band of the initial VLF / LF frequency band frequency domain division setting are unknown, the initial time domain related segment length is set based on empirical values.

[0093] Furthermore, the signal synthesis and processing module is specifically used for:

[0094] Calculate the short-time Fourier transform of each VLF / LF signal in the k-th iteration;

[0095] Frequency domain correlation processing is performed on each VLF / LF signal and the synthesized signal:

[0096]

[0097]

[0098] in This indicates that the i-th sub-band and j-th path obtained in the k-th iteration is centered at τ and has a duration of T. i k The result of the short-time Fourier transform of the signal, where f is the independent variable of the short-time Fourier transform, and T is the signal. i (k) This represents the time-domain segment length of the i-th sub-band signal in the k-th iteration; This indicates that the i-th sub-band obtained in the (k-1)-th iteration is centered at τ and has a duration of T. i k-1 The frequency domain synthesized signal, T i k-1 This represents the time-domain segment length of the i-th sub-band signal in the (k-1)-th iteration; F represents the frequency division region of the i-th sub-band in the (k-1)-th iteration; i (k-1) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the (k-1)-th iteration; All are process variables; Represents the integral variable;

[0099] For the frequency domain synthesis parameters of the k-th iteration, difference compensation is performed, and the frequency domain data after parameter difference compensation is calculated according to the following formula.

[0100]

[0101] Calculate the frequency domain data of the synthesized signal in the k-th iteration.

[0102]

[0103] Will It is passed to the signal distribution information detection module.

[0104] Furthermore, the signal distribution information detection module is specifically used for:

[0105] If the i-th sub-band flag S ≠ '0', then the k-th iteration is calculated according to the following formula. Relevant phase spectrum information

[0106]

[0107] in Indicates taking The imaginary part, Indicates taking The real part;

[0108] right Perform linear fitting to obtain the fitted linear function.

[0109]

[0110] in It is the slope of the fitted linear function. It is the intercept of the fitted linear function;

[0111] calculate and To determine the error between the two, set a signal detection threshold μ and compare the error with the threshold.

[0112] The signal frequency domain distribution information is updated. If the error value is less than the threshold, the sub-frequency band division remains unchanged. If the error value is greater than the threshold, the i-th sub-frequency band in the (k-1)-th iteration is evenly divided into two new sub-frequency bands, thus obtaining the sub-frequency band division of the corresponding frequency region in the k-th iteration.

[0113] Update the corresponding processing parameters based on the obtained updated sub-band allocation scheme, and then update the... T i (k) F i (k) The signal is transmitted to the signal synthesis and processing module, where This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the k-th iteration.

[0114] Compared with the prior art, the present invention has the following advantages:

[0115] ① Compared with existing methods, the method and system proposed in this invention have better adaptability to scenarios where multiple signals exist within the processing frequency band.

[0116] ② Applying this invention to the reception and processing of VLF / LF communication signals can continuously optimize the time-frequency segmentation scheme during iterative processing, better adapting to the frequency domain distribution and reception environment of VLF / LF band signals.

[0117] ③The method and system proposed in this invention fully transmit and utilize the information and signals extracted from the two processing stages of signal detection and signal synthesis, thus achieving better processing performance. Attached Figure Description

[0118] Figure 1 The flowchart shows a multi-antenna joint received signal detection method based on large number decision.

[0119] Figure 2 This is a flowchart of a multi-antenna joint received signal detection method based on signal synthesis;

[0120] Figure 3 This is a flowchart of a VLF / LF signal receiving and processing method according to an embodiment of the present invention;

[0121] Figure 4 This is a flowchart illustrating the signal frequency domain distribution information update process according to an embodiment of the present invention.

[0122] Figure 5 This is a schematic diagram of a VLF / LF signal receiving and processing system architecture according to an embodiment of the present invention. Detailed Implementation

[0123] For ease of understanding, the following explanations are provided for some of the terms used in the specific embodiments of this invention:

[0124] LF (Low Frequency): Low frequency.

[0125] VLF (Very Low Frequency): Extremely low frequency.

[0126] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments:

[0127] First, define the following parameter expression:

[0128] N: Number of VLF / LF signal receiving channels;

[0129] In the k-th iteration, the frequency division region of the i-th sub-band has a sub-band flag bit 'S' with values ​​including "0 / 1 / #", where S = '1' indicates that there is a signal, S = '0' indicates that there is no signal, and S = '#' indicates that no judgment is made yet.

[0130] T i (k) : The length of the time-domain segment of the i-th sub-frequency band signal in the k-th iteration;

[0131] F i (k) : The length of the frequency domain correlation segment of the i-th sub-band signal in the k-th iteration;

[0132] The i-th sub-band and j-th path obtained in the k-th iteration are centered at τ and have a duration of T. i k The result of the short-time Fourier transform of the signal; where f is the independent variable of the short-time Fourier transform;

[0133] The i-th sub-band obtained in the k-th iteration is centered at τ and has a duration of T. i k The frequency domain synthesized signal.

[0134] The basic scheme of a typical embodiment of the VLF / LF signal receiving and processing method of the present invention is as follows: Figure 3 As shown, Figure 3 The implementation plans for each processing stage are as follows:

[0135] ① Multi-channel sampling

[0136] The received multi-channel VLF / LF signals are subjected to direct sampling transformation across the entire frequency band from 3kHz to 300kHz, with a typical sampling rate of no less than 1MHz.

[0137] ② Initial processing parameter settings

[0138] The initial VLF / LF band frequency domain division, initial frequency domain related segment length, and initial time domain related segment length are set, and these parameters are passed to processing stages ③ and ④. A typical implementation scheme is as follows.

[0139] (1) A typical implementation scheme for setting the "initial VLF / LF band frequency domain division" is as follows:

[0140] If the signal frequency domain distribution within the VLF / LF band for receiving and processing is known, the initial VLF / LF band frequency domain division is set based on the signal frequency domain distribution information. For example, if there are N signals within the receiving and processing band (whether these N signals appear in the received samples is unknown, but their frequency domain distribution is known), the corresponding frequency domain division is set based on the lowest and highest frequencies of each signal.

[0141] If the number of signals and their frequency distribution within the VLF / LF band are unknown, the initial band division can be performed based on the signal reception characteristics of the VLF / LF band. A typical division scheme is as follows:

[0142]

[0143] If the frequency domain distribution information of the signal within the receiving and processing frequency band is known, then the known noise sub-frequency band flag is set to S = '0', and the initial flags of the remaining sub-frequency bands are all set to S = '#'.

[0144] (2) A typical implementation scheme for setting the "initial frequency domain correlation segment length" is as follows:

[0145] If the frequency domain distribution of the signal within the VLF / LF band is known, then If the number of signals and frequency domain distribution within the VLF / LF bands are unknown, then set F... i (0) =1.

[0146] (3) A typical implementation scheme for setting the "initial time domain segment length" is as follows:

[0147] If there is a known frequency region without effective signal (i.e. a frequency region with only noise) in the sub-band division scheme in (1), then noise samples collected by multiple receiving channels in that frequency region can be extracted, the variance characteristics of the noise samples can be statistically analyzed, and the initial time-domain correlation segment length of each sub-band can be set according to the time interval when the noise variance characteristics show significant differences.

[0148] If the frequency regions without effective signals in each sub-band divided in (1) are unknown, the initial time-domain correlation segment length can be set based on empirical values, with a typical value of 300ms.

[0149] ③ Multi-channel VLF / LF band signal synthesis

[0150] A typical implementation scheme is as follows: the received multi-channel VLF / LF full-band signals are integrated and processed.

[0151] Step 31: Calculate the short-time Fourier transform of each VLF / LF signal in the k-th iteration;

[0152] Based on the (k-1)th VLF / LF band frequency domain division Time-domain segment length {T i (k-1) Short-time Fourier transform is performed on each of the VLF / LF signals sampled from the multi-channel receiver to obtain the transformed frequency domain data.

[0153] Step 32: Perform frequency domain correlation processing on each VLF / LF signal and the synthesized signal;

[0154] The typical implementation scheme for correlation processing of frequency domain data of the i-th sub-band in the k-th iteration is as follows.

[0155]

[0156]

[0157] It is worth noting that, It's just a process variable, for further representation It's just a process variable, for further representation This indicates the integration variable, used to distinguish it from f;

[0158] Step 33: Compensation for differences in frequency domain synthesis parameters in the k-th iteration;

[0159] Calculate the frequency domain data after parameter difference compensation:

[0160]

[0161] Step 34: Calculate the frequency domain data of the synthesized signal in the k-th iteration.

[0162]

[0163] Step 35: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require the Passed to the processing stage ④.

[0164] ④ Detection of VLF / LF band signal distribution information

[0165] A typical implementation plan is as follows:

[0166] Step 41: Calculate the k-th iteration Relevant phase spectrum information

[0167] If the i-th sub-band flag S ≠ '0', then calculate the relevant phase spectrum information of the sub-band in the k-th iteration:

[0168]

[0169] in Indicates taking The imaginary part, Indicates taking The real part;

[0170] Step 42: Analyze the relevant phase spectrum information in the k-th iteration. Perform linear fitting;

[0171] Based on the relevant phase spectrum information in the k-th iteration Perform linear fitting to obtain the fitted linear function.

[0172]

[0173] in, It is the slope of the fitted linear function. It is the intercept of the fitted linear function;

[0174] Step 43: Calculation and judgment of correlation phase spectrum fitting error;

[0175] Calculate the correlation phase spectrum information in the k-th iteration The error between the linear fitting result and the error can be measured using different methods such as absolute error and mean square error. This example provides a typical scheme for measuring absolute error.

[0176]

[0177] in, For the calculation area The length.

[0178] Set the signal detection threshold μ and compare the error with the threshold value.

[0179] Step 44: Update signal frequency domain distribution information;

[0180] This example provides a typical update scheme, and its processing flow is as follows: Figure 4 As shown.

[0181] The typical implementation scheme for sub-band splitting is as follows:

[0182]

[0183] If the error value is less than the threshold, the sub-band division remains unchanged; if the error value is greater than the threshold, the i-th sub-band in the (k-1)-th iteration is evenly divided into two new sub-bands, resulting in the sub-band division of the frequency region corresponding to the k-th iteration.

[0184] Step 45: Process parameter updates and information transmission

[0185] Based on the updated sub-band allocation scheme obtained in step 44, the corresponding processing parameters are updated. A typical scheme is as follows:

[0186] The time-domain correlation segment length of each sub-band can be updated according to the detection results of the signal distribution, using the implementation scheme in processing step ② (3).

[0187] The typical update scheme for the frequency domain correlation segment length of each sub-band is as follows:

[0188] And the updated T i (k) F i (k) The data is then transferred to the processing stage ③.

[0189] ⑤ Convergence judgment and signal information output

[0190] Step 1: Convergence check;

[0191] The iteration loop terminates when the convergence condition is met between two consecutive iterations. A typical convergence criterion is as follows:

[0192]

[0193] Step 2: Signal information output;

[0194] The combined signal samples of each sub-frequency band of the existing signal are output after inverse Fourier transform.

[0195] Output the frequency domain distribution information of the received and processed frequency band signal.

[0196] Based on the above embodiments, such as Figure 5 As shown, another aspect of the present invention also proposes a VLF / LF signal receiving and processing system, comprising:

[0197] The multi-channel sampling module is used to perform multi-channel sampling on the received multiple VLF / LF signals;

[0198] The initial processing parameter setting module is used to set the initial processing parameters, which include the initial VLF / LF frequency band frequency domain division, the initial frequency domain related segment length, and the initial time domain related segment length.

[0199] The signal synthesis and processing module is used to perform integrated synthesis processing on multiple received VLF / LF full-band signals.

[0200] The signal distribution information detection module is used to detect the signal distribution information of the VLF / LF frequency band;

[0201] The convergence judgment and signal information output module is used to iteratively execute the integrated signal synthesis processing module and the signal distribution information detection module until the convergence condition is met. It outputs the synthesized signal samples of each sub-frequency band of the existing signal after inverse Fourier transform, and outputs the frequency domain distribution information of the received and processed frequency band signal obtained by detection.

[0202] Furthermore, the multi-channel sampling module is specifically used for:

[0203] The received multi-channel VLF / LF signals are subjected to direct sampling transformation across the entire frequency band of [3KHz~300KHz], with a sampling rate of not less than 1MHz.

[0204] Furthermore, the initial VLF / LF band frequency domain division settings include:

[0205] If the frequency domain distribution of the signal within the VLF / LF band is known, the initial VLF / LF band frequency domain division is set according to the signal frequency domain distribution information.

[0206] If the number of signals and frequency domain distribution information within the VLF / LF band are unknown, the initial band division is performed based on the signal reception characteristics of the VLF / LF band.

[0207] If the frequency domain distribution information of the signal within the receiving and processing frequency band is known, then the known noise sub-frequency band flag is set to S = '0', and the initial flags of the remaining sub-frequency bands are all set to S = '#'. S = '0' indicates no signal, and S = '#' indicates no judgment is made for the time being.

[0208] Furthermore, the setting of the initial frequency domain correlation segment length includes:

[0209] If the frequency domain distribution of the signal within the VLF / LF band is known, then Where F i (0) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the initial stage. This represents the frequency division region of the i-th sub-band in the initial stage;

[0210] If the number of signals and frequency domain distribution information within the VLF / LF bands to be received and processed are unknown, then set F i (0) =1.

[0211] Furthermore, the setting of the initial time-domain correlation segment length includes:

[0212] If there is a known frequency region without effective signal in the initial VLF / LF band frequency domain division setting, then the noise samples collected by multiple receiving channels in the frequency region are extracted, the variance characteristics of the noise samples are statistically analyzed, and the initial time domain correlation segment length of each sub-band is set according to the time interval when the noise variance characteristics show significant differences.

[0213] If the frequency regions without effective signals in each sub-band of the initial VLF / LF frequency band frequency domain division setting are unknown, the initial time domain related segment length is set based on empirical values.

[0214] Furthermore, the signal synthesis and processing module is specifically used for:

[0215] Calculate the short-time Fourier transform of each VLF / LF signal in the k-th iteration;

[0216] Frequency domain correlation processing is performed on each VLF / LF signal and the synthesized signal:

[0217]

[0218]

[0219] in This indicates that the i-th sub-band and j-th path obtained in the k-th iteration is centered at τ and has a duration of T. i k The result of the short-time Fourier transform of the signal, where f is the independent variable of the short-time Fourier transform, and T is the signal. i (k) This represents the time-domain segment length of the i-th sub-band signal in the k-th iteration; This indicates that the i-th sub-band obtained in the (k-1)-th iteration is centered at τ and has a duration of T. i k-1 The frequency domain synthesized signal, T i k-1 This represents the time-domain segment length of the i-th sub-band signal in the (k-1)-th iteration; F represents the frequency division region of the i-th sub-band in the (k-1)-th iteration; i (k-1) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the (k-1)-th iteration; All are process variables; Represents the integral variable;

[0220] For the frequency domain synthesis parameters of the k-th iteration, difference compensation is performed, and the frequency domain data after parameter difference compensation is calculated according to the following formula.

[0221]

[0222] Calculate the frequency domain data of the synthesized signal in the k-th iteration.

[0223]

[0224] Will It is passed to the signal distribution information detection module.

[0225] Furthermore, the signal distribution information detection module is specifically used for:

[0226] If the i-th sub-band flag S ≠ '0', then the k-th iteration is calculated according to the following formula. Relevant phase spectrum information

[0227]

[0228] in Indicates taking The imaginary part, Indicates taking The real part;

[0229] right Perform linear fitting to obtain the fitted linear function.

[0230]

[0231] in It is the slope of the fitted linear function. It is the intercept of the fitted linear function;

[0232] calculate and To determine the error between the two, set a signal detection threshold μ and compare the error with the threshold.

[0233] The signal frequency domain distribution information is updated. If the error value is less than the threshold, the sub-frequency band division remains unchanged. If the error value is greater than the threshold, the i-th sub-frequency band in the (k-1)-th iteration is evenly divided into two new sub-frequency bands, thus obtaining the sub-frequency band division of the corresponding frequency region in the k-th iteration.

[0234] Update the corresponding processing parameters based on the obtained updated sub-band allocation scheme, and then update the... T i (k) F i (k) The signal is transmitted to the signal synthesis and processing module, where This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the k-th iteration.

[0235] In summary, compared to existing methods, the method and system proposed in this invention are more adaptable to scenarios involving multiple signals within a processing frequency band. Applying this invention to VLF / LF communication signal reception processing allows for continuous optimization of the time-frequency segmentation scheme during iterative processing, better adapting to the frequency domain distribution of VLF / LF band signals and the receiving environment. The method and system proposed in this invention fully transmit and utilize the information and signals extracted from the signal detection and signal synthesis processing stages, thus achieving superior processing performance.

[0236] The above description is only a preferred embodiment of the present invention. It should be noted that those skilled in the art can make several improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A VLF / LF signal receiving and processing method, characterized in that, include: Step 1: Perform multi-channel sampling on the received multiple VLF / LF signals; Step 2: Set initial processing parameters, including initial VLF / LF frequency band frequency domain division, initial frequency domain related segment length, and initial time domain related segment length; Step 3: Perform integrated processing on the received multi-channel VLF / LF full-band signals; Step 4: Detect the signal distribution information of the VLF / LF band; Step 5: Iteratively execute steps 3 and 4 until the convergence condition is met. Output the synthesized signal samples of each sub-band containing the signal after inverse Fourier transform, and output the frequency domain distribution information of the received and processed frequency band signal obtained from the detection. Step 4 includes: Step 41: If the i-th sub-band flag S ≠ '0', then calculate the process variables in the k-th iteration according to the following formula. Relevant phase spectrum information in Indicates taking The imaginary part, Indicates taking The real part of , j represents the j-th path, f is the independent variable of the short-time Fourier transform, and τ is the center of the i-th sub-band; Step 42: [Regarding...] Perform linear fitting to obtain the fitted linear function. in It is the slope of the fitted linear function. It is the intercept of the fitted linear function; Step 43: Calculation To determine the error between the two, set a signal detection threshold μ and compare the error with the threshold. Step 44: Update the signal frequency domain distribution information. If the error value is less than the threshold, the sub-frequency band division remains unchanged. If the error value is greater than the threshold, the i-th sub-band in the (k-1)-th iteration is evenly divided into two new sub-bands, resulting in the sub-band division of the frequency region corresponding to the k-th iteration. Step 45: Update the corresponding processing parameters based on the updated sub-band allocation scheme obtained in Step 44, and then update the parameters accordingly. Pass to step 3, where T represents the frequency division region of the i-th sub-band in the k-th iteration. i (k) This represents the time-domain segment length of the i-th sub-band signal in the k-th iteration. F i (k) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the k-th iteration.

2. The VLF / LF signal receiving and processing method according to claim 1, characterized in that, Step 1 includes: The received multi-channel VLF / LF signals are subjected to direct sampling transformation across the entire frequency band of [3KHz~300KHz], with a sampling rate of not less than 1MHz.

3. The VLF / LF signal receiving and processing method according to claim 1, characterized in that, The initial VLF / LF band frequency domain division settings include: If the frequency domain distribution of the signal within the VLF / LF band is known, the initial VLF / LF band frequency domain division is set according to the signal frequency domain distribution information. If the number of signals and frequency domain distribution information within the VLF / LF band are unknown, the initial band division is performed based on the signal reception characteristics of the VLF / LF band. If the frequency domain distribution information of the signal within the receiving and processing frequency band is known, then the known noise sub-frequency band flag is set to S = '0', and the initial flags of the remaining sub-frequency bands are all set to S = '#'. S = '0' indicates no signal, and S = '#' indicates no judgment is made for the time being.

4. The VLF / LF signal receiving and processing method according to claim 1, characterized in that, The setting of the initial frequency domain correlation segment length includes: If the frequency domain distribution of the signal within the VLF / LF band is known, then Where F i (0) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the initial stage. This represents the frequency division region of the i-th sub-band in the initial stage; If the number of signals and frequency domain distribution information within the VLF / LF bands to be received and processed are unknown, then set F i (0) =1.

5. The VLF / LF signal receiving and processing method according to claim 1, characterized in that, The setting of the initial time-domain correlation segment length includes: If there is a known frequency region without effective signal in the initial VLF / LF band frequency domain division setting, then the noise samples collected by multiple receiving channels in the frequency region are extracted, the variance characteristics of the noise samples are statistically analyzed, and the initial time domain correlation segment length of each sub-band is set according to the time interval when the noise variance characteristics show significant differences. If the frequency regions without effective signals in each sub-band of the initial VLF / LF frequency band frequency domain division setting are unknown, the initial time domain related segment length is set based on empirical values.

6. The VLF / LF signal receiving and processing method according to claim 1, characterized in that, Step 3 includes: Step 31: Calculate the short-time Fourier transform of each VLF / LF signal in the k-th iteration; Step 32: Perform frequency domain correlation processing on each VLF / LF signal and the synthesized signal: in This indicates that the i-th sub-band and j-th path obtained in the k-th iteration is centered at τ and has a duration of T. i (k) The result of the short-time Fourier transform of the signal, where f is the independent variable of the short-time Fourier transform, and T is the signal. i (k) This represents the time-domain segment length of the i-th sub-band signal in the k-th iteration; This indicates that the i-th sub-band obtained in the (k-1)-th iteration is centered at τ and has a duration of T. i (k-1) The frequency domain synthesized signal, T i (k-1) This represents the time-domain segment length of the i-th sub-band signal in the (k-1)-th iteration; F represents the frequency division region of the i-th sub-band in the (k-1)-th iteration; i (k-1) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the (k-1)-th iteration; All are process variables; Represents the integral variable; Step 33: Perform difference compensation on the frequency domain synthesis parameters of the k-th iteration, and calculate the frequency domain data after parameter difference compensation according to the following formula. Step 34: Calculate the frequency domain data of the synthesized signal in the k-th iteration. Where N is the number of VLF / LF signal receiving channels; Step 35: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require the Pass it to step 4.

7. A VLF / LF signal receiving and processing system, characterized in that, include: The multi-channel sampling module is used to perform multi-channel sampling on the received multiple VLF / LF signals; The initial processing parameter setting module is used to set the initial processing parameters, which include the initial VLF / LF frequency band frequency domain division, the initial frequency domain related segment length, and the initial time domain related segment length. The signal synthesis and processing module is used to perform integrated synthesis processing on the received multiple VLF / LF full-band signals. The signal distribution information detection module is used to detect the signal distribution information of the VLF / LF frequency band; The convergence judgment and signal information output module is used to iteratively execute the integrated signal synthesis processing module and the signal distribution information detection module until the convergence condition is met. It outputs the synthesized signal samples of each sub-frequency band of the existing signal after inverse Fourier transform, and outputs the frequency domain distribution information of the received and processed frequency band signal obtained by detection. The signal distribution information detection module is specifically used for: Step 41: If the i-th sub-band flag S ≠ '0', then calculate the process variables in the k-th iteration according to the following formula. Relevant phase spectrum information in Indicates taking The imaginary part, Indicates taking The real part of , j represents the j-th path, f is the independent variable of the short-time Fourier transform, and τ is the center of the i-th sub-band; Step 42: [Regarding...] Perform linear fitting to obtain the fitted linear function. in It is the slope of the fitted linear function. It is the intercept of the fitted linear function; Step 43: Calculation To determine the error between the two, set a signal detection threshold μ and compare the error with the threshold. Step 44: Update the signal frequency domain distribution information. If the error value is less than the threshold, the sub-frequency band division remains unchanged. If the error value is greater than the threshold, the i-th sub-band in the (k-1)-th iteration is evenly divided into two new sub-bands, resulting in the sub-band division of the frequency region corresponding to the k-th iteration. Step 45: Update the corresponding processing parameters based on the updated sub-band allocation scheme obtained in Step 44, and then update the parameters accordingly. The signal is transmitted to the signal synthesis and processing module, where T represents the frequency division region of the i-th sub-band in the k-th iteration. i (k) This represents the time-domain segment length of the i-th sub-band signal in the k-th iteration. F i (k) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the k-th iteration.

8. The VLF / LF signal receiving and processing system according to claim 7, characterized in that, The signal synthesis and processing module is specifically used for: Calculate the short-time Fourier transform of each VLF / LF signal in the k-th iteration; Frequency domain correlation processing is performed on each VLF / LF signal and the synthesized signal: in This indicates that the i-th sub-band and j-th path obtained in the k-th iteration is centered at τ and has a duration of T. i (k) The result of the short-time Fourier transform of the signal, where f is the independent variable of the short-time Fourier transform, and T is the signal. i (k) This represents the time-domain segment length of the i-th sub-band signal in the k-th iteration; This indicates that the i-th sub-band obtained in the (k-1)-th iteration is centered at τ and has a duration of T. i (k-1) The frequency domain synthesized signal, T i (k-1) This represents the time-domain segment length of the i-th sub-band signal in the (k-1)-th iteration; F represents the frequency division region of the i-th sub-band in the (k-1)-th iteration; i (k-1) This represents the length of the frequency domain correlation segment of the i-th sub-band signal in the (k-1)-th iteration; All are process variables; Represents the integral variable; For the frequency domain synthesis parameters of the k-th iteration, difference compensation is performed, and the frequency domain data after parameter difference compensation is calculated according to the following formula. Calculate the frequency domain data of the synthesized signal in the k-th iteration. Where N is the number of VLF / LF signal receiving channels; Will It is passed to the signal distribution information detection module.

Images