Near space intermediate frequency radar correlation function noise reduction processing method

Abstract

The invention provides a near space intermediate frequency radar correlation function noise reduction processing method. The processing method comprises the following steps: calculating a correlation function according to a complex time sequence received by an antenna; eliminating a zero delay point in the correlation function, and fitting data of n points on the left and right of the zero delay point to obtain a fitting function; calculating a noise factor based on the fitting function; and performing noise reduction processing on the correlation function by using the noise factor. The method can solve the problem of inaccuracy of the correlation function of the MF radar receiving signal.

Description

Technical Field

[0001] The present invention belongs to the technical field of radio detection in the space environment, and specifically relates to a method for reducing noise of the correlation function of a near-space medium-frequency radar. Background Art

[0002] The near space generally refers to the aerospace transition region within the range of 20 - 120 km. Its environmental parameters mainly include wind field, density, temperature, etc. When the above parameters are violently disturbed, it will have a negative impact on the flight trajectory, state and life of near-space aircraft. At present, the main means for detecting the wind field in the near space of 60 - 90 km include rocket payloads, meteor radars, medium-frequency radars, etc. Among them, rocket payloads are costly and have insufficient time and space resolution. As one of the important conventional detection means in the D region, the medium-frequency (MF) radar can not only detect the wind field in the D region, but also obtain the electron density, and has been widely deployed globally for conventional observations. The medium-frequency radar mainly uses the Fresnel partial reflection principle, that is, the electromagnetic waves O-wave and X-wave emitted by the radar are partially reflected by inhomogeneous clumps. According to the full correlation relationship between the O-wave and X-wave signals, the full correlation coefficient between the transmitted power and the received power is calculated, and the atmospheric wind field is further inverted using the theory of the full correlation analysis method. The use of this method is mainly restricted by two aspects. On the one hand, it is the interaction relationship of scattering, reflection, diffuse reflection, etc. of inhomogeneous clumps in the middle atmosphere on electromagnetic waves. On the other hand, it is the influence of background environmental noise on the signal propagation path. The first aspect belongs to the characteristics of the natural atmosphere itself, while the second aspect belongs to the influence of external interference on signal propagation.

[0003] In the application of the full correlation (Full Correlation Function, FCA) method of medium-frequency radars, the correlation function is the most basic calculation quantity. The random fluctuations of scatterers, the boundedness of radar targets, etc. will have a significant impact on the echo sequence, making the correlation function have an uneven spike at zero delay. In addition, atmospheric background noise, power supply noise, man-made noise, receiver noise, and galactic background noise, etc. will also greatly affect the correlation function, making it show some uneven burrs. As the input variable of the FCA method, the accuracy of the correlation function directly affects the correctness of the inverted wind field components. Therefore, it is necessary to propose a method for suppressing noise, and no relevant prior art has been found on this content. Summary of the Invention

[0004] The purpose of the present invention aims to solve at least one of the problems existing in the prior art.

[0005] For this reason, the present invention provides a method for reducing noise of the correlation function of a near-space medium-frequency radar.

[0006] The technical solution of the present invention is as follows:

[0007] According to one aspect, there is provided a method for reducing noise in the correlation function of a near-space medium-frequency radar, and the processing includes:

[0008] Calculating the correlation function according to the complex time series received by the antenna;

[0009] Removing the zero-delay point in the correlation function, and taking the data of n points on the left and right of the zero-delay point for fitting to obtain a fitting function;

[0010] Calculating the noise factor based on the fitting function;

[0011] Using the noise factor to perform noise reduction processing on the correlation function.

[0012] Furthermore, the correlation function is composed of an autocorrelation function and a cross-correlation function, and both the autocorrelation function and the cross-correlation function are processed as follows: removing the zero-delay point in the correlation function, and taking the data of n points on the left and right of the zero-delay point for fitting to obtain a fitting function.

[0013] Furthermore, the cross-correlation function is calculated according to the complex time series received by the antenna by the following formula:

[0014]

[0015] where ρ ij (τ) represents the cross-correlation function; f i (t) and f j (t) are both complex time series, the subscripts i and j both represent the receiving antenna numbers, t is the echo time, and f i (t) = s i (t) + n i (t), s i (t) is the signal, and n i (t) is the noise. Similarly, s j (t) and n j (t) are respectively the signal and noise that make up f j (t); where n i (t) = n ci (t) + n ui (t), n i (t) is further divided into noise n ci (t) that is also correlated between different antenna pairs and noise n ui (t) that is not correlated between different antenna pairs. Similarly, n j (t) = n cj (t) + n uj (t), and the superscript * represents the conjugate.

[0016] Furthermore, the autocorrelation function is calculated according to the complex time series received by the antenna by the following formula:

[0017]

[0018] Among them, ρ(τ) represents the autocorrelation function.

[0019] Furthermore, when performing fitting, quadratic polynomial fitting is adopted.

[0020] Furthermore, the fitting point number n is determined by the following method:

[0021] If the gap between the signal-to-noise ratio SNR calculated by the function fitted according to the selected n fitting points and the true signal-to-noise ratio SNR meets the design threshold, then the number of the selected n fitting points meets the requirements.

[0022] Furthermore, the noise factor is solved by the following method for both the fitting function corresponding to the autocorrelation function and the fitting function corresponding to the cross-correlation function:

[0023] Let the value of the fitted function at the zero-delay point multiplied by the noise factor be the value 1 of the noise-free autocorrelation function at the zero-delay point. Therefore, the noise factor is 1 divided by the value of the fitted function at the zero-delay point.

[0024] Furthermore, after denoising the correlation function using the noise factor, the estimation of the cross-correlation function without noise influence is:

[0025]

[0026] Furthermore, after denoising the correlation function using the noise factor, the estimation of the cross-correlation function without noise influence is:

[0027]

[0028] According to another aspect, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the above-mentioned processing method is implemented.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] Using the method for reducing noise in the solution of the present invention, the change characteristics of the radar received signal at different signal-to-noise ratios can be very well matched, and are infinitely close to the signal quality without noise. The method of the present invention has been successfully applied to the radar system and greatly improves the observation accuracy of the radar. For example, before denoising, the signal amplitude is only about 0.4, and after denoising, the signal amplitude can reach 0.9, and the signal quality is improved by 125%. Description of the Drawings

[0031] The accompanying drawings included are used to provide a further understanding of the embodiments of the present invention, which form a part of the specification, illustrate the embodiments of the present invention, and, together with the written description, explain the principles of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0032] Figure 1 is the autocorrelation function of the sine function sin(t) at different signal-to-noise ratios;

[0033] Figure 2 is the fitting error caused by too many fitting points;

[0034] wherein, (a) 32-point quadratic polynomial fitting; (b) 128-point quadratic polynomial fitting;

[0035] Figure 3 is the estimation error of different signal-to-noise ratios using second-order polynomial fitting at different fitting points;

[0036] Figure 4 is the delay τ 0.5 error;

[0037] Figure 5 is the principle of fitting calculation;

[0038] Figure 6 is the noise reduction effect of the autocorrelation function;

[0039] Figure 7 is the influence of noise on the cross-correlation function of pairs;

[0040] Figure 8 is the noise reduction effect of the cross-correlation function. Detailed implementation manners

[0041] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, rather than all, of the embodiments of the present invention. The following description of at least one exemplary embodiment is actually only illustrative and in no way limits the present invention and its application or use. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0042] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0043] Unless otherwise specifically stated, the relative arrangements of components and steps, numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present invention. At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn in actual proportional relationships. Technologies, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the said technologies, methods, and devices should be regarded as part of the authorized specification. In all the examples shown and discussed herein, any specific values should be construed as merely exemplary and not as limitations. Therefore, other examples of the exemplary embodiments may have different values. It should be noted that: like reference numerals and letters denote like items in the following drawings, and thus, once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0044] The technical problem to be solved by the present invention is the inaccuracy of the correlation function of the MF radar received signal. Based on the analysis of the MF radar detection principle and the influence of noise on the correlation function, therefore, an embodiment of the present invention proposes a method for denoising the correlation function of a near-space intermediate-frequency radar. Specifically:

[0045] In an embodiment of the present invention, there is provided a method for denoising the correlation function of a near-space intermediate-frequency radar, and the processing includes:

[0046] Step 1: Calculate the correlation function according to the complex time series received by the antenna;

[0047] Step 2: Eliminate the zero-delay point in the correlation function, and take the data of n points on the left and right of the zero-delay point for fitting to obtain a fitting function;

[0048] Step 3: Calculate the noise factor based on the fitting function;

[0049] Step 4: Use the noise factor to perform denoising processing on the correlation function.

[0050] In the embodiment of the present invention, the correlation function is composed of an autocorrelation function and a cross-correlation function, and both the autocorrelation function and the cross-correlation function are processed as follows: Eliminate the zero-delay point in the correlation function, and take the data of n points on the left and right of the zero-delay point for fitting to obtain a fitting function.

[0051] That is, in the embodiments of the present invention, noise reduction processing needs to be performed on both the autocorrelation function and the cross-correlation function.

[0052] In the above embodiment, the cross-correlation function is calculated according to the following formula based on the complex time series received by the antenna:

[0053]

[0054] where ρ ij (τ) represents the cross-correlation function; f i (t) and f j (t) are both complex time series, the subscripts i and j both represent the receiving antenna numbers, t is the echo time, and f i (t) = s i (t) + n i (t), s i (t) is the signal, and n i (t) is the noise. Similarly, s j (t) and n j (t) are respectively the signal and the noise that make up f j (t); where n i (t) = n ci (t) + n ui (t), n i (t) is further divided into the noise n ci (t) that is correlated between different antenna pairs and the noise n ui (t) that is not correlated between different antenna pairs. Similarly, n j (t) = n cj (t) + n uj (t), and the superscript * represents the conjugate.

[0055] Specifically, assume that the complex time series received by antenna A i is f i (t), and this sequence is composed of the signal s i (t) and the noise n i (t)

[0056] f i (t) = s i (t) + n i (t) (1)

[0057] Since the signal and the noise are uncorrelated, their statistical average value is 0

[0058]

[0059] In the above formula, the subscript i or j represents the receiving antenna number, t represents the echo time, and τ represents the time delay, which is the same hereinafter. Among them, the noise can be further divided into two forms. One is the noise n ci (t) that is also correlated between different antenna pairs. This kind of noise is usually composed of atmospheric background noise, power supply noise, and man-made noise. The other is the noise n ui (t) that is not correlated between different antenna pairs. This kind of noise includes receiver noise and galactic background noise. Thus, we get

[0060] n i (t) = n ci (t) + n ui (t) (3)

[0061] Since whether it is n ci (t) or n ui (t), their respective statistical averages over time are 0. Thus, the following rule can be obtained

[0062]

[0063]

[0064]

[0065]

[0066]

[0067] The cross-correlation function of the complex time series f i (t) and f j (t) is

[0068]

[0069] Expanding the complex time series f i (t) and f j (t) as a combination of signals and noise, the cross-correlation function can be rewritten as

[0070]

[0071] Expanding its numerator, we can get

[0072]

[0073] Among them and are 0. After these two terms are eliminated, the remaining terms can be further expanded as

[0074]

[0075] By canceling out the last three terms on the right side of the equation, the numerator can be rewritten as

[0076]

[0077] Expanding the denominator in a similar way gives

[0078]

[0079] Simplifying it to

[0080]

[0081] Finally, the expression for the cross-correlation function can be derived as follows

[0082]

[0083] In the above embodiment, the autocorrelation function is calculated from the complex time series received by the antenna according to the following formula:

[0084]

[0085] where ρ(τ) represents the autocorrelation function.

[0086] Specifically, the autocorrelation function is expanded and simplified using a method similar to that for obtaining the cross-correlation function. The autocorrelation function of the time series f i (t) is defined as

[0087]

[0088] If f i (t) is expanded as a combination of a signal and noise, the autocorrelation function can be rewritten as

[0089]

[0090] Similarly, expanding the numerator gives

[0091]

[0092] The second and third terms on the right side of the equal sign can be eliminated, thus simplifying the numerator to

[0093]

[0094] Then, continuing to expand the denominator further gives

[0095]

[0096] The second and third terms on the right side of the equal sign can be eliminated, thus simplifying the denominator to

[0097] D(ρ(τ)) = <|s i (t)| 2 > + <|n i (t)| 2 > (22)

[0098] Finally, the autocorrelation function is simplified to

[0099]

[0100] Analyzing the above equation, it can be clearly seen that when the noise intensity increases, the value at non-zero delay points (τ≠0) becomes smaller and smaller, while at the zero-delay point τ = 0, it remains a fixed value of 1, which will result in an obvious "peak" at the zero-delay point.

[0101] Preferably, when performing fitting, quadratic polynomial fitting is adopted.

[0102] Specifically, the zero-delay point in the correlation function is removed, and the data of n points on the left and right of the zero-delay point are taken for Gaussian fitting or quadratic polynomial fitting. However, it should be noted that before applying this noise reduction algorithm to radar data, the number of fitting points and the specific selection of the fitting method are very important processing details, which need to be considered from two aspects: the fitting method and the number of points. Among them, since the correlation function is even symmetric, even-order polynomial fitting and cubic spline fitting are considered. After calculation, the error of quadratic polynomial fitting is the smallest, so quadratic polynomial is selected as the fitting method.

[0103] In the above embodiment, the fitting point number n is determined by the following method:

[0104] If the difference between the signal-to-noise ratio SNR calculated by the function fitted with the selected n fitting points and the true signal-to-noise ratio SNR meets the design threshold, then the number of the selected n fitting points meets the requirements.

[0105] That is to say, the number of fitting points has a great influence on the quadratic polynomial fitting result. When the number of fitting points is appropriate, the shape of the fitted function is very close to the autocorrelation function. However, if the number of fitting points is too large, the contour of the obtained correlation function values is no longer the contour of the quadratic function, so there will be a large error. If the number of fitting points is too small, the information provided for fitting is too little, which will also lead to a large error in fitting.

[0106] Since the true value of SNR in the simulation is known, the embodiments of the present invention consider using the difference between the SNR calculated from the results fitted with different fitting points and the true SNR (i.e., the SNR estimation error) as the performance criterion to investigate the influence of the number of fitting points on this performance. Obviously, the smaller the SNR estimation error, the better the performance.

[0107] In the above embodiment, the fitting function corresponding to the autocorrelation function and the fitting function corresponding to the cross-correlation function both calculate the noise factor in the following manner:

[0108] The value of the fitted function at the zero delay point is multiplied by the noise factor to obtain the value of the noise-free autocorrelation function at the zero delay point, which is 1. Therefore, the noise factor is 1 divided by the value of the fitted function at the zero delay point.

[0109] That is, after obtaining the noise factor, each point of the correlation function is divided by the noise factor to perform noise reduction processing on each point.

[0110] Furthermore, after the noise factor is used to perform noise reduction processing on the correlation function, the estimation of the cross-correlation function without noise influence is:

[0111]

[0112] Furthermore, after the noise factor is used to perform noise reduction processing on the correlation function, the estimation of the cross-correlation function without noise influence is:

[0113]

[0114] Specifically, the value of the fitted function at zero time delay is multiplied by the noise factor to obtain the value of the noise-free autocorrelation function at zero time delay (ie, 1), so the noise factor is 1 divided by the value of the fitted function at zero time delay.

[0115] Since the FCA method is based on the assumption that the correlation function is not affected by noise, it is necessary to try to reduce the noise of the correlation function to fit the assumption of FCA.

[0116] First, the zero-delay point of the autocorrelation function is replaced by a "noise-free" estimate, and we get

[0117]

[0118] Divide the value by a factor

[0119]

[0120] So we get a noise-free estimate of the autocorrelation function

[0121]

[0122] Similar to the autocorrelation function, the value of the cross-correlation function at the 0 delay point is rewritten as

[0123]

[0124] Divide this value by the factor An estimate of the cross-correlation function without the influence of noise can be obtained as

[0125]

[0126] As shown in Table 1, the effects of noise on the autocorrelation and cross-correlation functions and the corresponding noise reduction methods are summarized

[0127] Table 1

[0128]

[0129] In summary, the noise reduction processing method using the solution of the present invention can make the variation characteristics of the radar received signal at different signal-to-noise ratios very well matched, and infinitely close to the signal quality without noise. The method of the present invention has been successfully applied to the radar system and greatly improves the observation accuracy of the radar. For example, before the noise reduction processing, the signal amplitude is only about 0.4, and after the noise reduction processing, the signal amplitude can reach 0.9, and the signal quality is improved by 125%.

[0130] To analyze the performance of the noise reduction method of the embodiments of the present invention, the following will be demonstrated by numerical simulation examples. To simplify the analysis, a sine signal is used as the transmitted signal.

[0131] First, a 200-point sine signal s(t)=sin(t), (t = 0, 0.1,..., 19.9) is generated. For the case of additive white Gaussian noise, the signal-to-noise ratio (SNR) is defined as

[0132]

[0133] where σ 2 is the variance of the additive white Gaussian noise, and different SNR conditions of the receiver can be simulated by adjustment. Calculate the corresponding autocorrelation function of the sine signal in the additive white Gaussian noise with different SNR, and the results are as Figure 1 shown.

[0134] Furthermore, carry out fitting and calculate the noise factor and the signal-to-noise ratio. Specifically as follows:

[0135] (1) Take a 100-point sine signal, and add 0 dB of additive white Gaussian noise to the credit signal under the defined SNR to simulate the received signal;

[0136] (2) Conduct 2000 independent repeated experiments on the received signal to obtain 2000 received signal samples with additive white Gaussian noise;

[0137] (3) Perform second-order polynomial fitting on the received signal samples. The fitting is carried out using 23 different numbers of fitting points, namely 4, 6... 40, and the average error between the SNR calculated from the fitting results and the true value is statistically obtained.

[0138] (4) Change the SNR to 5dB, 10dB, 15dB, 20dB and repeat the above process to obtain a set of estimated values of SNR under different SNRs. Statistically obtain the mean square error of the SNR estimation.

[0139]

[0140] Compare the variation of the estimation errors of the quadratic polynomial fitting under all different SNRs with respect to the time delay. It can be seen that the more the number of fitting points, the greater the SNR estimation error. In addition, at the same fitting point and fitting method, as the noise intensity increases, the SNR estimation error also increases. Considering the characteristics of the intermediate frequency radar echo data, the details of the fitting method used for noise reduction processing of the correlation function in the MF radar received signal are as follows: Second-order polynomial fitting is adopted, and the number of fitting points is taken as 8. The influence of the number of fitting points on the quadratic polynomial fitting results is as Figure 2 shown. The estimation errors of different signal-to-noise ratios using the second-order polynomial fitting under different numbers of fitting points are as Figure 3 shown.

[0141] Obviously, noise will affect the shape of the autocorrelation function, causing it to deform. Considering that the FCA method needs to take the time delay τ 0.5 corresponding to the function value of the autocorrelation function falling to 0.5 as a relevant parameter for estimating the elliptical parameters. Obviously, τ 0.5 in the deformed autocorrelation function will be directly affected. The following Figure 4 intuitively shows this influence. It can be clearly seen that the error of τ 0.5 will definitely be transmitted to the estimation of the elliptical parameters. Therefore, noise reduction processing must be carried out on the autocorrelation function.

[0142] Next, the fitting process and the estimation method of SNR derived during the fitting process will be demonstrated. Here, the autocorrelation function of a sine function with 15dB additive Gaussian white noise added is taken as an example. The specific operations in this part are shown in the drawing. The original autocorrelation function is the curve with triangle symbols in the figure. Take a total of 8 data points on both sides of the 0 time delay point of the autocorrelation function for fitting calculation. These points are marked as squares in the figure. Then, use a second-order polynomial to fit these eight points, and the fitting result is a parabola, as Figure 5 shown. Take the vertex value ρ nf (0) of the parabola and calculate the corresponding derivative as the noise factor F i .

[0143] Note that the SNR is only applicable to theoretical analysis, and the specific value of SNR is usually difficult to calculate directly in practical engineering applications. However, based on ρ nf determined by the fitting method, an estimated value of SNR can be further derived as

[0144]

[0145] Next, the noise factor will be used to restore the autocorrelation function affected by noise. The restoration method is to multiply the function value at each point of the autocorrelation function by the noise factor F i . As shown in the figure above, it can be seen that at most function values, they approach the noise-free function values. Therefore, the restoration effect of the noise factor is relatively satisfactory.

[0146] The noise reduction process of the cross-correlation function is as follows:

[0147] (1) s i (t) = sin(t), s j (t) = cos(t), (t = 0, 0.1,..., 19.9), calculate the cross-correlation functions of s i (t) and s j (t) under additive white Gaussian noise with different SNRs;

[0148] (2) The autocorrelation functions of the two sine signals can be calculated by the above method to obtain two noise factors F i and F j , then the noise factor used for noise reduction of the cross-correlation function is as Figure 7 shown;

[0149] (3) After multiplying each data point in the above figure by the corresponding noise factor, the noise reduction result is as follows Figure 8 shown.

[0150] It can be seen that similar to the noise reduction effect of the autocorrelation function in Figure 6 , after the cross-correlation function is processed by fitting for noise reduction, the cross-correlation function values at most sampling points under each SNR fluctuate around the noise-free cross-correlation function value with small fluctuations. Although it is impossible to make the cross-correlation function values of the sampling points under noise converge to the cross-correlation function in the noise-free case, the subsequent inversion accuracy requirements of the FCA method are already satisfied in engineering applications.

[0151] The features described and / or illustrated for one embodiment above can be used in the same or similar manner in one or more other embodiments, and / or combined with the features in other embodiments or used to replace the features in other embodiments.

[0152] It should be emphasized that the term "comprising / including" as used herein refers to the presence of features, whole units, steps or components, but does not exclude the presence or addition of one or more other features, whole units, steps, components or combinations thereof.

[0153] The above method of the present invention can be implemented by hardware or by a combination of hardware and software. The present invention relates to such a computer-readable program that, when executed by a logic component, can enable the logic component to implement the device or component described above, or enable the logic component to implement the various methods or steps described above. The present invention also relates to a storage medium for storing the above program, such as a hard disk, a magnetic disk, an optical disk, a DVD, a flash memory, etc.

[0154] Many features and advantages of these embodiments are apparent from this detailed description, and thus the appended claims are intended to cover all such features and advantages of these embodiments that fall within their true spirit and scope. In addition, since many modifications and changes are readily envisioned by those skilled in the art, the embodiments of the present invention are not to be limited to the exact structures and operations illustrated and described, but may cover all suitable modifications and equivalents that fall within their scope.

[0155] The parts not detailed in the present invention are well-known techniques to those skilled in the art.

Claims

1. A method for reducing noise of the correlation function of a near-space medium-frequency radar, characterized in that, The processing method includes: Calculating a correlation function according to the complex time series received by the antenna; Removing the zero-delay point in the correlation function, and taking the data of n points on the left and right of the zero-delay point for fitting to obtain a fitting function; Solving the noise factor based on the fitting function; Performing noise reduction processing on the correlation function by using the noise factor.

2. A method for noise reduction processing of the correlation function of a near-space medium-frequency radar according to claim 1, characterized in that, The correlation function is composed of an autocorrelation function and a cross-correlation function. Both the autocorrelation function and the cross-correlation function are processed as follows: removing the zero-delay point in the correlation function, and taking the data of n points on the left and right of the zero-delay point for fitting to obtain a fitting function.

3. A method for noise reduction processing of the correlation function of a near-space medium-frequency radar according to claim 2, characterized in that, The cross-correlation function is calculated according to the complex time series received by the antenna by the following formula: where, ρ ij (τ) represents the cross-correlation function; f i (t) and f j (t) are both complex time series, the subscripts i and j both represent the receiving antenna numbers, t is the echo time, f i (t) = s i (t) + n i (t), s i (t) is the signal, n i (t) is the noise, similarly s j (t) and n j (t) are respectively the signal and the noise that constitute f j (t); where, n i (t) = n ci (t) + n ui (t), n i (t) is further subdivided into the noise n ci (t) that is also correlated between different antenna pairs and the noise n ui (t) that is not correlated between different antenna pairs, similarly n j (t) = n cj (t) + n uj (t), the superscript * represents the conjugate.

4. A method for reducing noise of correlation functions of a near-space medium-frequency radar according to claim 3, characterized in that The autocorrelation function is calculated according to the complex time series received by the antenna by the following formula: where ρ(τ) represents the autocorrelation function.

5. A method for noise reduction processing of a near-space medium-frequency radar correlation function according to claim 3 or 4, characterized in that When performing fitting, quadratic polynomial fitting is adopted.

6. A method for noise reduction processing of correlation functions of a near-space medium-frequency radar according to claim 3 or 4, characterized in that, The number of fitting points n is determined by the following method: If the gap between the signal-to-noise ratio SNR calculated by the function fitted according to the selected n fitting points and the true signal-to-noise ratio SNR meets the design threshold, the number of the selected n fitting points meets the requirements.

7. A method for noise reduction processing of the correlation function of a near-space medium-frequency radar according to claim 6, characterized in that, The noise factor is solved for the fitting function corresponding to the autocorrelation function and the fitting function corresponding to the cross-correlation function by the following method: Let the value of the fitted function at the zero-delay point multiplied by the noise factor be the value 1 of the noise-free autocorrelation function at the zero-delay point. Therefore, the noise factor is 1 divided by the value of the fitted function at the zero-delay point.

8. A method for noise reduction processing of the correlation function of a near-space medium-frequency radar according to claim 7, characterized in that, After performing noise reduction processing on the correlation function by using the noise factor, the estimate of the cross-correlation function without noise influence is:

9. A method for noise reduction processing of the correlation function of a near-space medium-frequency radar according to claim 8, characterized in that, After performing noise reduction processing on the correlation function by using the noise factor, the estimate of the cross-correlation function without noise influence is:

10. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, the processing method described in claims 1-9 is implemented.

Images