Data processing method and system in short-wave communication system

Abstract

The invention relates to the technical field of short-wave selection, in particular to a data processing method and system in a short-wave communication system, and the method comprises the following steps: constructing a compensation matrix based on time-varying channel response, carrying out the multipath alignment of a received signal, and generating a preliminary equalization signal; sequentially carrying out time domain impulse noise suppression and frequency domain background noise correction on the preliminary equalization signal to generate a noise reduction signal; and according to a weighted evaluation result of the signal distortion degree and the noise residual quantity, dynamically adjusting a compensation matrix parameter and a noise reduction intensity coefficient to form closed-loop control. According to the invention, in a dynamic channel with a low signal-to-noise ratio, synchronous suppression of multipath interference and noise is realized, and signal distortion and redundancy overhead are avoided at the same time.

Description

Technical Field

[0001] The present application relates to the technical field of shortwave selection, and in particular to a data processing method and system in a shortwave communication system. Background Art

[0002] Shortwave communication is widely used in maritime, emergency communications and other fields due to its long-distance transmission capability. However, shortwave channels are affected by time-varying reflections from the ionosphere and suffer from severe dynamic multipath interference. At the same time, due to the long transmission distance and complex environment, the signal is often polluted by strong noise. Traditional methods usually use step-by-step processing: first suppress multipath interference through an equalizer, and then reduce noise through a filter. However, this discrete processing has the following defects:

[0003] 1. Noise amplification problem: When the signal-to-noise ratio (SNR<5dB) is low, the multipath equalizer will amplify the noise power, causing a sharp increase in the bit error rate (BER).

[0004] 2. Parameter mismatch problem: The parameters of multipath separation and noise suppression are designed independently and cannot adapt to the time-varying characteristics of the channel. Especially in the scenario of sudden interference, the convergence speed is insufficient.

[0005] 3. Signal distortion accumulation: The cascade processing of time domain filtering and frequency domain noise reduction may introduce phase shift and spectrum leakage, resulting in loss of signal details.

[0006] How to achieve simultaneous suppression of multipath interference and noise in a low signal-to-noise ratio dynamic channel while avoiding signal distortion and redundant overhead remains a technical challenge that urgently needs to be overcome in the field of shortwave communications. Summary of the invention

[0007] In order to achieve synchronous suppression of multipath interference and noise in a low signal-to-noise ratio dynamic channel while avoiding signal distortion and redundant overhead, the present application provides a data processing method and system in a shortwave communication system.

[0008] In a first aspect, the present application provides a data processing method and system in a shortwave communication system, which adopts the following technical solution: A data processing method and system in a shortwave communication system, comprising the following steps: Based on the time-varying channel response, a compensation matrix is ​​constructed to perform multipath alignment on the received signal and generate a preliminary equalized signal; Sequentially performing time domain impulse noise suppression and frequency domain background noise correction on the preliminary equalized signal to generate a noise reduction signal;

[0009] According to the weighted evaluation results of signal distortion and residual noise, the compensation matrix parameters and noise reduction intensity coefficient are dynamically adjusted to form a closed-loop control.

[0010] In one of the embodiments: The time-varying channel response is constructed in real time through multipath parameters. Among them, the constructed time-varying channel response matrix is: ; In the formula, k is the number of multipaths, is the time-varying attenuation coefficient of the k-th path, is the Doppler frequency shift, is the delay spread, is the multipath delay.

[0011] In one of the embodiments: The compensation matrix is: ; In the formula, is the conjugate transpose of the time-varying channel matrix, is the estimated value of the noise power, is the estimated value of the signal power, is the noise suppression factor.

[0012] In one of the embodiments: The time-domain impulse noise suppression adopts symmetric sliding window filtering, and its formula is: ; In the formula, is the time-domain filtered signal, is the preliminary equalization signal, and M is the window length; Among them, the preliminary equalization signal is the product of the compensation matrix parameter and the received signal.

[0013] In one of the embodiments: The formula for obtaining the noise-reduced signal through the frequency-domain background noise correction is: ; In the formula, represents the amplitude of the f-th frequency point, is the spectrum of the time-domain filtered signal, is the noise reduction intensity coefficient. Among them, the spectrum of the time-domain filtered signal is obtained by performing a Fourier transform on the time-domain filtered signal.

[0014] In one of the embodiments: The weighted evaluation result is achieved by constructing a cost function, and the cost function is: ; In the formula, is the weight factor, is the ideal time-domain signal, is the processed estimated signal, is the frequency-domain noise residue, is the signal distortion degree, is the noise residue; among them, the processed estimated signal is obtained by performing an inverse Fourier transform on the noise-reduced signal.

[0015] In one embodiment: the weight factor is dynamically adjusted, and its formula is: ; In the formula, is the signal-to-noise ratio adjustment factor, is the signal-to-noise ratio.

[0016] In one embodiment: the noise reduction intensity coefficient is dynamically adjusted by satisfying the condition of λ = 1 - β between the noise reduction intensity coefficient and the weight factor.

[0017] In one embodiment: the formula for dynamically adjusting the compensation matrix parameter is: ; In the formula, is the compensation matrix update step size, and the compensation matrix update step size satisfies with the noise reduction intensity coefficient.

[0018] In a second aspect, the present application provides a data processing method and system in a short-wave communication system, adopting the following technical solutions:

[0019] A data processing system in a short-wave communication system includes: A dynamic multipath separation module that constructs a compensation matrix based on the time-varying channel response, performs multipath alignment on the received signal, and generates a preliminary equalized signal; A time-frequency domain joint noise reduction module that sequentially performs time-domain impulse noise suppression and frequency-domain background noise correction on the preliminary equalized signal to generate a noise-reduced signal; A joint parameter optimization module that dynamically adjusts the compensation matrix parameter and the noise reduction intensity coefficient according to the weighted evaluation result of the signal distortion degree and the noise residue, forming a closed-loop control. Description of the Drawings

[0020] Figure 1 is a step schematic diagram of the data processing method in the short-wave communication system of this embodiment. Detailed Embodiments

[0021] The following further describes the present application in detail with reference to the drawings.

[0022] Embodiment 1: A data processing method in a short-wave communication system, as Figure 1 shown, includes the following steps:

[0023] S200. Construct a compensation matrix based on the time-varying channel response, perform multipath alignment on the received signal, and generate a preliminary equalized signal.

[0024] When the signal encounters obstacles in the environment such as buildings, vehicles, and trees, reflection, refraction, and diffraction will occur, forming multiple propagation paths. The propagation distances of different propagation paths are different, resulting in different arrival times of the signal at the receiving end, that is, time delay. Therefore, the receiving end will receive the signals of multiple paths simultaneously to form signal superposition, and these signals interfere with each other due to time delay and phase differences.

[0025] In this regard, in this step, the time-varying channel response is constructed in real time through multipath parameters to dynamically track the changes in multipath parameters caused by ionospheric reflection and improve the compensation accuracy. Among them, the constructed time-varying channel response matrix is: ; In the formula, k is the number of multipaths, is the time-varying attenuation coefficient of the k-th path, and its value range is 0.1~0.8, is the Doppler frequency shift, is the delay spread, is the multipath delay.

[0026] Then, a compensation matrix is constructed through the time-varying channel response. By estimating in real time, the system can track the multipath changes and dynamically adjust the compensation matrix. Through the coupling design, the bit error rate can be basically reduced from of the traditional method when SNR = 5dB to .

[0027] Among them, the constructed compensation matrix is: ; In the formula, is the conjugate transpose of the time-varying channel matrix, which is used to perform phase alignment and amplitude compensation on the received signal to align the signals of different paths in the time domain. For example, if a certain reflection path causes the signal phase to rotate by 30°, will generate a rotation of -30° to correct it.

[0028] By summing the squares of the moduli of all elements of the matrix, the energy overload of the multipath signal is prevented. For example, when the total energy is too large, the amplification factor is automatically reduced to avoid signal distortion.

[0029] is the noise power estimate value, is the signal power estimate value, is the noise suppression factor, indicating the noise ratio.

[0030] In this compensation matrix, the numerator term is used to cancel the phase rotation of each path, and noise suppression is adopted in the denominator term to suppress noise and avoid being affected by Overamplification. For example When = 0.1 (low noise), the compensation matrix focuses on multipath compensation; while

[0031] When = 1 (high noise), the compensation matrix focuses on noise suppression. In this way, through the dynamic adjustment of the noise-signal power ratio in the denominator, the contradiction between "correcting multipath" and "not amplifying noise" in the traditional scheme is solved. In the formula, is the received signal.

[0032] S400. Perform time-domain impulse noise suppression and frequency-domain background noise correction on the preliminary equalization signal in sequence to generate a noise-reduced signal.

[0033] In step S200, after multipath compensation, the multipath interference can be reduced by more than 60%, but the signal is still mixed with burst noise (such as lightning pulses) and continuous white noise. Therefore, perform time-domain impulse noise suppression and frequency-domain background noise correction on the preliminary equalization signal in sequence for noise reduction processing, and form a multi-scale processing design by combining time-domain coarse processing and frequency-domain fine processing.

[0034] Time-domain impulse noise suppression uses symmetric sliding window filtering to smooth the burst noise, and its formula is:[[]] ; In the formula, is the time-domain filtered signal, is the preliminary equalization signal, and M is the window length.

[0035] Suppress white noise through frequency-domain background noise correction to obtain the noise-reduced signal, and its formula is specifically:[[]] ; In the formula, represents the amplitude of the f-th frequency point, is the spectrum of the time-domain filtered signal; is the noise reduction intensity coefficient, and the noise reduction intensity coefficient is dynamically adjusted according to the signal-to-noise ratio, and its range = 0.1~1.0.

[0036] Among them, the spectrum of the time-domain filtered signal is obtained by performing a Fourier transform on the time-domain filtered signal , that is:[[]] ; S600. Dynamically adjust the compensation matrix parameters and the noise reduction intensity coefficient according to the weighted evaluation results of the signal distortion degree and the noise residue amount to form a closed-loop control.

[0037] The weighted evaluation result is achieved by constructing a cost function, and the cost function is: ; In the formula, is the weight factor, is the ideal time-domain signal, is the processed estimated signal, is the frequency-domain noise residue.

[0038] Among them, the processed estimated signal is obtained by performing an inverse Fourier transform on the noise-reduced signal, that is: ; The calculation formula for the frequency-domain noise residue is: ; In the formula, is the spectrum of the ideal noise-free signal.

[0039] The acquisition of the spectrum of the ideal noise-free signal is divided into two cases: with known pilot signals and without pilots. If the system inserts pilots (training sequences), it is directly generated by the pilot symbols, and its formula is: ; In the formula, is the known time-domain pilot signal preset at the sending end.

[0040] If no pilots are used, the ideal spectrum is reconstructed from the demodulated signal, and the specific steps are: Perform a hard decision on the noise-reduced time-domain signal to obtain a symbol sequence; Remodulate the symbol sequence into the ideal time-domain signal ; Perform an FFT on to generate .

[0041] is the signal distortion degree, is the noise residue amount, and adaptive balance is achieved through the weight factor.

[0042] The weight factor is dynamically adjusted, and its formula is: ; In the formula, is the signal-to-noise ratio adjustment factor, is the signal-to-noise ratio.

[0043] Among them, the calculation formula for the signal-to-noise ratio is: ;

[0044] Dynamically adjust the noise reduction intensity coefficient by satisfying the condition of λ = 1 - β between the noise reduction intensity coefficient and the weight factor.

[0045] The formula for dynamically adjusting the compensation matrix parameters is: ; ; In the formula, is the update step size of the compensation matrix.

[0046] The update step size of the compensation matrix and the noise reduction intensity coefficient satisfy , so as to achieve dynamic coupling with the noise reduction intensity coefficient and balance the convergence speed and stability. Therefore, by adopting the design of dynamic step size, the number of convergence iterations can be achieved, and the bit error rates of low SNR and high SNR can be reduced.

[0047] This optimization algorithm adopts an improved gradient descent method, making the maximum number of iterations ≤ 10 times, and usually able to achieve the number of iterations ≤ 5 times, that is, usually reaching the optimal balance point after at most 5 iterations, making J reach the minimum value, greatly reducing the number of iterations and the computational complexity.

[0048] In another embodiment, the present application provides a data processing system in a short-wave communication system, and the system includes: A dynamic multipath separation module, which constructs a compensation matrix based on the time-varying channel response, performs multipath alignment on the received signal, and generates a preliminary equalized signal; A time-frequency domain joint noise reduction module, which sequentially performs time-domain impulse noise suppression and frequency-domain background noise correction on the preliminary equalized signal to generate a noise-reduced signal; A joint parameter optimization module, which dynamically adjusts the compensation matrix parameters and the noise reduction intensity coefficient according to the weighted evaluation results of the signal distortion degree and the noise residue amount to form a closed-loop control.

[0049] The embodiments of this specific implementation manner are all preferred embodiments of the present application, and do not limit the protection scope of the present application accordingly. Therefore, all equivalent changes made according to the structure, shape, and principle of the present application should be covered within the protection scope of the present application.

Claims

1. A data processing method in a shortwave communication system, characterized in that: The following steps are involved: Based on the time-varying channel response, a compensation matrix is ​​constructed to perform multipath alignment on the received signal and generate a preliminary equalized signal; Sequentially performing time domain impulse noise suppression and frequency domain background noise correction on the preliminary equalized signal to generate a noise reduction signal; According to the weighted evaluation results of signal distortion and residual noise, the compensation matrix parameters and noise reduction intensity coefficient are dynamically adjusted to form a closed-loop control.

2. The data processing method in a shortwave communication system according to claim 1, characterized in that: The time-varying channel response is constructed in real time through multipath parameters, wherein the constructed time-varying channel response matrix is: ； Where k is the number of multipaths, is the time-varying attenuation coefficient of the kth path, is the Doppler shift, For delay extension, is the multipath delay.

3. The data processing method in a shortwave communication system according to claim 2, characterized in that: The compensation matrix is: ； In the formula, is the conjugate transpose of the time-varying channel matrix, is the noise power estimate, is the estimated signal power, is the noise suppression factor.

4. The data processing method in a shortwave communication system according to claim 1, characterized in that: The time domain impulse noise suppression adopts symmetrical sliding window filtering, and its formula is: ； In the formula, is the time domain filtered signal, is the preliminary equalized signal, M is the window length; The preliminary equalization signal is the product of the compensation matrix parameter and the received signal.

5. The data processing method in a shortwave communication system according to claim 4, characterized in that: The formula for obtaining the noise reduction signal by correcting the frequency domain background noise is: ； In the formula, represents the amplitude of the fth frequency point, is the spectrum of the time-domain filtered signal, is the noise reduction strength coefficient, wherein the spectrum of the time domain filtered signal is obtained by performing Fourier transform on the time domain filtered signal.

6. The data processing method in a shortwave communication system according to claim 4, characterized in that: The weighted evaluation result is achieved by constructing a cost function, which is: ； In the formula, is the weight factor, is an ideal time domain signal, is the estimated signal after processing, is the residual noise in the frequency domain, is the signal distortion, is the noise residual; wherein the processed estimated signal is obtained by performing inverse Fourier transform on the noise reduction signal.

7. The data processing method in a shortwave communication system according to claim 6, characterized in that: The weight factor is dynamically adjusted, and its formula is: ； In the formula, is the signal-to-noise ratio adjustment factor, is the signal-to-noise ratio.

8. The data processing method in a shortwave communication system according to claim 7, characterized in that: The noise reduction intensity coefficient is dynamically adjusted by satisfying the condition of λ=1-β between the noise reduction intensity coefficient and the weight factor.

9. The data processing method in a shortwave communication system according to claim 1, characterized in that: The formula for dynamically adjusting the compensation matrix parameters is: ； In the formula, is the learning rate, and the learning rate and the denoising intensity coefficient satisfy .

10. A data processing system in a shortwave communication system, characterized in that: include: The dynamic multipath separation module builds a compensation matrix based on the time-varying channel response, performs multipath alignment on the received signal, and generates a preliminary equalized signal; A time-frequency domain joint noise reduction module performs time-domain impulse noise suppression and frequency-domain background noise correction on the preliminary equalized signal in sequence to generate a noise reduction signal; The joint parameter optimization module dynamically adjusts the compensation matrix parameters and the noise reduction intensity coefficient according to the weighted evaluation results of the signal distortion and the noise residual to form a closed-loop control.

Images