Real-time suppression method for narrowband interference of satellite navigation based on power inversion

Abstract

The application relates to a satellite navigation narrow-band interference real-time suppression method based on power inversion. The method comprises the following steps: segmenting and windowing a continuous satellite navigation digital intermediate frequency signal, converting the signal to a frequency domain through FFT, realizing power inversion through unit circle projection to attenuate narrow-band interference, determining an effective frequency point index range according to a signal center frequency and a nominal bandwidth, zeroing out out-of-band spectral lines, recovering a time domain signal through IFFT, and finally recombining and outputting through overlapping post-processing. The method does not require prior information and threshold judgment throughout the process, can blindly and adaptively suppress interference, solves the problem of signal-to-noise ratio deterioration under a wide sampling bandwidth, has strong real-time performance, small signal distortion, and is suitable for multiple navigation systems.

Description

A Real-Time Suppression Method for Narrowband Interference in Satellite Navigation Based on Power Inversion Technical Field

[0001] This application relates to the field of satellite navigation signal processing technology, and in particular to a real-time method for suppressing narrowband interference in satellite navigation based on power inversion. Background Technology

[0002] Satellite navigation signals, transmitted from satellites to ground receivers, have extremely low power and are often submerged in thermal noise, making them highly susceptible to intentional or unintentional radio interference. Narrowband interference, due to its concentrated energy, ease of generation, and effectiveness, is one of the most threatening forms of interference to GNSS receivers. This type of interference can originate from communication equipment in adjacent frequency bands, harmonic radiation, or deliberate jammers, severely degrading the receiver's correlation peaks, causing tracking loop lock-up, and even completely paralyzing the receiver. The core of time-domain adaptive filtering methods is to construct an adaptive filter in the time domain and adjust its coefficients in real time to create nulls at the interference frequencies. Mainstream methods include: FIR adaptive filtering: simple in structure, requires iterative updates of filter coefficients, has a trade-off between convergence speed and steady-state performance, and may fail when interference changes rapidly. IIR adaptive notch filtering: can deeply suppress specific single-tone or multi-tone interference.

[0003] However, its performance heavily relies on the accurate estimation and tracking of interference frequencies. In dynamically changing interference scenarios, filters may exhibit transient responses due to convergence lag, leading to additional distortion in the navigation signal. Furthermore, these methods typically require a complex trade-off between suppressing interference and avoiding signal distortion, and parameter adjustments are not flexible enough. Frequency domain elimination methods convert the signal to the frequency domain for processing, and have received widespread attention due to their ability to intuitively identify and process interference spectral lines. The general process is as follows: perform a Fast Fourier Transform (FFT) on the received signal, identify and eliminate (zero out) spectral lines considered interference in the frequency domain, and finally recover the time-domain signal through an Inverse Fast Fourier Transform (IFFT). The core challenge and inherent limitation of this method lies in how to set a universal and accurate power threshold to distinguish between "interference spectral lines" and "signal plus noise spectral lines," which is the core challenge of this technology. 1) Too high a threshold: In pursuit of "clean" interference elimination, a large number of weak signal spectral lines carrying navigation information may be mistakenly zeroed out, resulting in excessive signal energy loss and a severe decrease in the equivalent carrier-to-noise ratio, directly affecting acquisition sensitivity and tracking accuracy. 2) Threshold too low: In order to avoid signal loss, some interference energy will remain, the suppression effect will be greatly reduced, and the receiver performance will still be damaged.

[0004] To set reasonable thresholds, existing methods typically require real-time estimation of the noise floor power of the received signal. However, in the presence of strong interference, noise power estimation itself is challenging, and estimation errors can directly lead to threshold failure. Furthermore, some improved algorithms attempt to introduce more complex statistical models or machine learning methods for interference detection, but this further increases the complexity of the algorithms and the requirements for prior knowledge of the scene.

[0005] The complete FFT->noise estimation->threshold calculation and decision->spectral line removal->IFFT processing chain is quite long. In particular, noise estimation and complex decision logic introduce non-negligible computational delays and uncertainties, making it difficult to meet the real-time processing requirements of highly dynamic users (such as aerospace vehicles and autonomous vehicles) for determinism and extremely low latency.

[0006] To improve multipath resistance and multi-frequency compatibility, modern high-precision GNSS receivers often employ sampling rates much larger than the bandwidth of the navigation signal itself (e.g., 16MHz or even higher sampling rates for GPS L1C / A code signals with a bandwidth of approximately 2MHz). In such "wide sampling bandwidth" systems, traditional frequency domain rejection methods face new challenges: the power of out-of-band noise spectral lines is significantly lower than in-band noise due to front-end analog filtering. Simply applying a uniform threshold removes high-power in-band interference and signal spectral lines, while retaining low-power out-of-band noise spectral lines. In subsequent normalization or reconstruction processes, this effectively amplifies out-of-band noise, causing the overall signal-to-noise ratio of the output signal to decrease rather than increase.

[0007] In summary, existing narrowband interference suppression techniques, especially mainstream frequency domain elimination methods, suffer from inherent contradictions in terms of adaptive threshold setting, reliance on prior knowledge, real-time processing, and adaptability to wide sampling bandwidths. Therefore, developing a novel narrowband interference suppression method that is independent of prior information and threshold decisions, possesses deterministic and low-latency processing capabilities, and can adapt to wide sampling bandwidth systems has become an urgent technical requirement for improving the survivability and performance of GNSS receivers in complex electromagnetic environments. Summary of the Invention

[0008] Therefore, it is necessary to provide a real-time satellite navigation narrowband interference suppression method based on power inversion that can solve the problems of difficulty in setting thresholds, reliance on noise estimation, and poor real-time performance in existing frequency domain elimination methods, while avoiding in-band signal loss caused by excessive receiver sampling bandwidth.

[0009] A real-time method for suppressing narrowband interference in satellite navigation based on power inversion, the method comprising:

[0010] The continuously input satellite navigation digital intermediate frequency signal stream is segmented and divided into continuous data blocks of fixed length. A window function is applied to each data block to obtain a windowed data block.

[0011] Perform a Fast Fourier Transform on each windowed data block to convert the time-domain signal to the frequency domain and obtain the corresponding complex spectrum; perform a unit circle projection operation on each complex spectrum in the frequency domain to obtain a frequency domain data block with inverted power.

[0012] Based on the center frequency and nominal bandwidth of the satellite navigation digital intermediate frequency signal, the corresponding effective frequency point index range is determined in the digital frequency domain. Spectral lines outside the effective frequency point index range are forcibly set to zero to obtain the optimized frequency domain data block.

[0013] Perform an inverse fast Fourier transform on the optimized frequency domain data block to obtain the recovered time domain signal;

[0014] The recovered time-domain signal is post-processed according to the overlap method, and the suppressed time-domain signal is reconstructed and output.

[0015] The aforementioned real-time suppression method for narrowband interference in satellite navigation based on power inversion first employs a power inversion operation using the unit circle projection of spectral lines. This requires no prior information or decision threshold. Through a blind adaptive mechanism that suppresses strong interference and supports weak interference, the amplitude of high-energy narrowband interference spectral lines is forcibly normalized to 1, significantly attenuating the interference energy. Simultaneously, the spectral shape of weak navigation signals and noise is relatively preserved, fundamentally avoiding performance degradation caused by inaccurate parameter estimation and significantly improving robustness. Secondly, by accurately calculating the effective frequency index range of the navigation signal, out-of-band spectral lines are forcibly set to zero, strictly limiting the processing gain within the effective bandwidth. This completely blocks the dilution effect of out-of-band noise, ensuring the output signal-to-noise ratio while suppressing interference. Furthermore, addressing the issues of long chain lengths and poor real-time performance in traditional methods, the entire process consists of segmented windowing, FFT, power inversion, out-of-band zeroing, IFFT, and overlapping recombination. All these are deterministic open-loop operations with no iteration or feedback loops, resulting in fixed and predictable processing latency. The simple structure facilitates parallel pipelined implementation on FPGAs / ASICs, meeting the real-time requirements of high-dynamic scenarios such as aerospace and autonomous driving. In addition, the soft suppression characteristic of power inversion, compared to traditional hard-decision elimination methods, can smoothly compress interference peaks, maximizing the preservation of the navigation signal spectrum shape, reducing signal distortion, and facilitating subsequent high-precision positioning, achieving a better balance between interference suppression and signal preservation. Simultaneously, it has fewer core parameters and configurable signal bandwidth, flexibly adapting to various navigation systems such as GPS, BDS, and Galileo, reducing engineering implementation and expansion costs, and comprehensively resolving the inherent contradictions of existing technologies in terms of adaptability, wide bandwidth adaptation, and real-time performance. Attached Figure Description

[0016] Figure 1 is a flowchart illustrating a real-time suppression method for narrowband interference in satellite navigation based on power inversion in one embodiment;

[0017] Figure 2 is a schematic diagram comparing the effects of narrowband interference suppression before and after the present application in one embodiment; wherein, (a) is the effect diagram before narrowband interference suppression by the present application, and (b) is the effect diagram after narrowband interference suppression by the present application;

[0018] Figure 3 shows the capture correlation peak results obtained before and after narrowband interference suppression according to this application in one embodiment; wherein, (a) is the capture correlation peak result obtained after narrowband interference suppression according to this application, and (b) is the capture correlation peak result obtained before narrowband interference suppression according to this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0020] In one embodiment, as shown in Figure 1, a real-time method for suppressing narrowband interference in satellite navigation based on power inversion is provided, including the following steps:

[0021] Step 102: The continuously input satellite navigation digital intermediate frequency signal stream is segmented and processed into continuous data blocks of fixed length. A window function is applied to each data block to obtain a windowed data block.

[0022] Satellite navigation digital intermediate frequency (IF) signal stream refers to the navigation signal received by the receiver and converted from analog to digital. It includes GNSS signals such as GPS, BeiDou, GLONASS, and Galileo, and is susceptible to narrowband interference. Segmentation processing divides the continuous signal into discrete data blocks of fixed length to facilitate subsequent FFT processing. The fixed length, where N is an integer power of 2, satisfies the FFT algorithm's input length requirement. Window functions are used to reduce spectral leakage caused by signal segmentation; Hanning or Hamming windows can be selected. Windowed data blocks are the result of multiplying the original data block by the window function point by point, preserving the core signal information while suppressing spectral leakage. This step lays the foundation for subsequent time-frequency transformation and interference suppression. Overlap processing (overlap preservation method or overlap addition method) ensures the continuity of the time-domain signal and avoids signal breakage after processing.

[0023] Step 104: Perform a fast Fourier transform on each windowed data block to convert the time-domain signal to the frequency domain and obtain the corresponding complex spectrum; perform a unit circle projection operation on each complex spectrum in the frequency domain to obtain a frequency domain data block with inverted power.

[0024] The Fast Fourier Transform (FFT) is an efficient algorithm for converting time-domain signals into frequency-domain signals, revealing the energy distribution of a signal at different frequencies. The complex spectrum contains amplitude and phase information for each frequency point, where amplitude reflects the energy intensity at that frequency and phase reflects the phase characteristics of the signal. The unit circle projection operation is the core innovation of this method. By dividing each complex spectral line by its own magnitude, the magnitude of all spectral lines after processing is made equal to 1, achieving power inversion—frequency points with stronger energy are suppressed more significantly, while weak navigation signals and noise are relatively preserved. This step eliminates the need for threshold decision and noise estimation, achieving blind adaptive interference suppression and completely eliminating dependence on prior information.

[0025] Step 106: Based on the center frequency and nominal bandwidth of the satellite navigation digital intermediate frequency signal, determine the corresponding effective frequency index range in the digital frequency domain, and force the spectral lines outside the effective frequency index range to zero to obtain the optimized frequency domain data block.

[0026] The center frequency is the carrier frequency of the navigation signal, and the nominal bandwidth is the frequency range of the navigation signal itself. The effective frequency index range is the set of spectral lines corresponding to the effective bandwidth of the navigation signal in the digital frequency domain, calculated using the center frequency and the nominal bandwidth. Zeroing out-of-band spectral lines forces all spectral lines outside the effective frequency band to zero; its core function is to solve the problem of out-of-band noise being relatively amplified under wide sampling bandwidth. This step ensures that the processing gain is concentrated in the useful signal frequency band, avoids out-of-band noise diluting the useful signal, and guarantees the carrier-to-noise ratio of the output signal.

[0027] Step 108: Perform an inverse fast Fourier transform on the optimized frequency domain data block to obtain the recovered time domain signal.

[0028] The Inverse Fast Fourier Transform (IFFT) is the inverse operation of the FFT, converting the frequency domain data, after interference suppression and out-of-band zeroing, back into a time domain signal. The recovered time domain signal retains the core information of the navigation signal, suppresses narrowband interference, and removes out-of-band noise; it still exists in the form of data blocks. This step completes the inverse time-frequency domain conversion, providing the time domain data foundation for subsequent signal reconstruction and output. The processing delay is fixed, meeting real-time requirements.

[0029] Step 110: Perform corresponding post-processing on the recovered time-domain signal according to the overlap method, reconstruct and output the suppressed time-domain signal.

[0030] The overlapping method is the overlap-save method or overlap-add method adopted in step 102, and the post-processing needs to correspond to it; the overlap-save method needs to remove the distorted part generated by cyclic convolution in the data block, and the overlap-add method needs to add the overlapping parts of adjacent data blocks; recombination and output is to splice discrete data blocks into a continuous time-domain signal stream; the time-domain signal that has been suppressed is a navigation signal in which narrowband interference is greatly attenuated and out-of-band noise is removed, and can be directly input into baseband processing modules such as acquisition and tracking of the receiver. This step ensures the continuity and integrity of the output signal and realizes real-time and continuous interference suppression processing.

[0031] The above real-time satellite navigation narrowband interference suppression method based on power inversion first uses the power inversion operation of spectral line unit circle projection, without any prior information and decision threshold. Through a blind adaptive mechanism of suppressing the strong and supporting the weak, the amplitude of the high-energy narrowband interference spectral line is forced to be normalized to 1, greatly attenuating the interference energy. At the same time, the spectral shapes of weak navigation signals and noise are relatively retained, avoiding performance degradation caused by inaccurate parameter estimation from the root, and significantly improving the robustness. Secondly, by accurately calculating the effective frequency point index range of the navigation signal, the out-of-band spectral lines are forced to be set to zero, strictly limiting the processing gain within the effective bandwidth, completely blocking the dilution effect of out-of-band noise, and ensuring the output signal-to-noise ratio while suppressing interference. Furthermore, aiming at the problems of long processing chain length and poor real-time performance of traditional methods, the entire process consists of segmented windowing, FFT, power inversion, out-of-band zeroing, IFFT, and overlap recombination, all of which are deterministic open-loop operations without iterative and feedback links. The processing delay is fixed and predictable, and the structure is simple and easy to implement in parallel pipelines of FPGA / ASIC, meeting the real-time requirements of high-dynamic scenarios such as aerospace vehicles and autonomous driving. In addition, compared with the traditional hard decision rejection method, the soft suppression characteristic of power inversion can smoothly compress the interference peak value, maintain the spectral shape of the navigation signal to the greatest extent, reduce signal distortion, and is more conducive to subsequent high-precision positioning, achieving a better balance between interference suppression and signal retention. At the same time, there are few core parameters and the signal bandwidth is configurable, which can flexibly adapt to various navigation systems such as GPS, BDS, Galileo, etc., reducing the engineering implementation and expansion costs, and comprehensively solving the internal contradictions of the existing technology in terms of adaptability, wide bandwidth adaptation, real-time performance, etc.

[0032] In one embodiment, the length N of the data block is an integer power of 2, the number of overlapping samples L between adjacent data blocks satisfies 0 < L < N, the window function is a Hanning window or a Hamming window, and the window function application formula is: , where is the i-th sample of the m-th original data block, is the value of the window function at the i-th position, is the windowed data block.

[0033] Specifically, the data block length N is chosen as an integer power of 2 to accommodate the efficient operation of the FFT algorithm and reduce computational complexity. The number of overlapping samples L between adjacent data blocks is set according to actual needs. The core purpose of overlap processing is to offset the attenuation effect of the window function on the edges of the data blocks, ensuring that the reconstructed time-domain signal is continuous and distortion-free. A balance is achieved between spectral leakage suppression and main lobe width, making it suitable for scenarios with high signal resolution requirements. When the window function is applied, it is multiplied point-by-point with the original data block, resulting in a smooth transition at the edges of the data blocks, effectively reducing spectral leakage caused by segmentation processing, and providing accurate spectral data for subsequent frequency domain interference suppression.

[0034] In one embodiment, a Fast Fourier Transform is performed on each windowed data block to convert the time-domain signal to the frequency domain, obtaining the corresponding complex spectrum, including:

[0035] Perform a Fast Fourier Transform on each windowed data block to convert the time-domain signal to the frequency domain, obtaining the corresponding complex spectrum.

[0036] ;

[0037] Where k is the spectral line index. It is a complex spectrum.

[0038] Specifically, after converting the time-domain signal to the frequency domain using FFT, narrowband interference manifests as high-amplitude spectral lines near specific frequency points, which contrast sharply with the low-amplitude spectral lines of navigation signals and noise. This provides a clear basis for interference localization in subsequent power inversion suppression. Transforming interference that is difficult to separate in the time domain into intuitively identifiable high-amplitude spectral lines in the frequency domain creates conditions for interference suppression.

[0039] In one embodiment, a unit circle projection operation is independently performed on each complex spectrum in the frequency domain to obtain a frequency domain data block with inverted power, including:

[0040] Perform a unit circle projection operation independently on each complex spectrum in the frequency domain to obtain a frequency domain data block with inverted power.

[0041] ;

[0042] in, It is a complex spectrum. It is a frequency domain data block after power inversion.

[0043] Specifically, this operation essentially involves projecting each spectral line onto a unit circle, meaning that regardless of its original amplitude, its modulus is always 1 after processing. For strong narrowband interference lines, its... The value is very large, and after dividing by its own magnitude, the amplitude is significantly attenuated (from a large number to 1). For weak navigation signals and noise spectra, its... The value is very small, so this operation changes its amplitude relatively little (from a decimal to 1), and may even be relatively boosted in some cases (when the original amplitude is much smaller than 1). The process achieves power inversion completely adaptively—the higher the energy of the frequency component, the more it is suppressed. It requires no prior or real-time estimation of noise power and no setting of any decision threshold, completely eliminating dependence on prior information. Processed spectrum. In this process, the peaks of interference are flattened, while the spectral shapes of the signal and noise are relatively preserved.

[0044] In one embodiment, the effective frequency point index range is determined in the digital frequency domain based on the center frequency and nominal bandwidth of the satellite navigation digital intermediate frequency signal, including:

[0045] Based on the center frequency of satellite navigation digital intermediate frequency signal and nominal bandwidth Determine the corresponding effective frequency index range in the digital frequency domain. ,in, Indicates the starting valid frequency point. This indicates the termination of the valid frequency point.

[0046] Specifically, the core function of determining the corresponding effective frequency index range in the digital frequency domain is to accurately define the frequency domain range of the useful signal, providing a basis for subsequent out-of-band spectral line zeroing and avoiding the dilution of the useful signal by out-of-band noise under wide sampling bandwidth.

[0047] In one embodiment, the calculation methods for the start effective frequency point and the end effective frequency point are as follows:

[0048] ;

[0049] ;

[0050] in, This is the rounding function. The lower limit frequency of the effective frequency band. Sampling frequency, The upper limit frequency of the effective frequency band. The number of points in the FFT, i.e., the length of the data block.

[0051] In one embodiment, the lower limit frequency and the upper limit frequency of the effective frequency band are calculated as follows:

[0052] ;

[0053] ;

[0054] in, The center frequency of the navigation signal. This is the nominal bandwidth of the navigation signal. Sampling frequency, This is the modulo operation.

[0055] Specifically, this calculation method takes into account scenarios with different intermediate frequency frequencies and sampling frequencies, ensuring the accuracy of the effective frequency band range and providing reliable input for the subsequent calculation of the effective frequency point index range.

[0056] In one embodiment, forcing spectral lines located outside the effective frequency index range to zero includes:

[0057] The process of forcibly setting spectral lines outside the effective frequency index range to zero is as follows:

[0058] ;

[0059] in, This refers to the frequency domain data after zeroing, i.e., the optimized frequency domain data block. It is a frequency domain data block after power inversion processing. It is the effective frequency index range of the navigation signal, and k is the spectral line index.

[0060] In one embodiment, an inverse fast Fourier transform is performed on the optimized frequency domain data block to obtain the recovered time domain signal, including:

[0061] Performing an inverse fast Fourier transform on the optimized frequency domain data block yields the recovered time domain signal.

[0062] ;

[0063] in, This refers to the frequency domain data after zeroing, i.e., the optimized frequency domain data block. The number of points in the FFT. is the index of the time-domain sample, and k is the spectral line index.

[0064] In one embodiment, the segmentation process employs either the overlap-preservation method or the overlap-addition method.

[0065] Specifically, the overlapping and preserving method works as follows: each data block has a length of N, adjacent data blocks overlap by L samples, after FFT processing, the first L distorted samples are discarded (due to circular convolution), and the last NL valid samples are retained. Finally, all valid samples are concatenated to obtain a continuous signal. The overlapping and adding method works as follows: each data block has a length of N, adjacent data blocks overlap by L samples, after FFT processing, the overlapping portions are added to offset the edge attenuation of the window function, ensuring signal continuity. Both methods can guarantee the integrity of the processed time-domain signal and avoid signal breakage. For example, using the overlapping and adding method with N=1024 and L=512, 512 new samples are read each time and concatenated with the 512 overlapping samples of the previous data block. After processing, the first 512 samples of the current data block are added to the last 512 samples of the previous data block to achieve continuous output. This overlapping processing method, combined with the window function, effectively suppresses spectral leakage and signal distortion, ensuring that the interference-suppressed signal can be directly used for baseband processing such as acquisition and tracking in the receiver.

[0066] In a specific embodiment, S1. Signal segmentation and windowing: The received zero-IF complex signal (I and Q channels) is The processing module processes data from... Read R=512 new samples from the previous data block and concatenate them with L=512 samples retained from the previous data block to form a new data block of length N=1024. , Apply a Hamming window to this data block. Obtain the windowed data block .

[0067] S2. Perform FFT (Time-Frequency Transform):

[0068] right Perform an FFT with N=1024 points to obtain its complex spectrum. ,in .

[0069] S3. Power inversion processing to suppress interference:

[0070] right Projecting each complex spectral line onto the unit circle:

[0071] ;

[0072] After this step, all (Except for points where the amplitude was originally 0). This means that the amplitudes of the spectral lines corresponding to the high-power 2MHz narrowband interference in the original spectrum are significantly attenuated (from a large number to 1); while the amplitudes of the spectral lines of the BeiDou B3I signal (lower power than noise) and broadband noise change relatively little.

[0073] S4. Beidou B3I signal out-of-band spectral lines return to zero:

[0074] Since it is a zero-IF signal, the effective frequency band of the BeiDou B3I signal is -10.23~10.23MHz. Calculate the corresponding effective frequency index range:

[0075] ;

[0076] ;

[0077] ;

[0078] ;

[0079] therefore:

[0080] ;

[0081] Execution reset to zero:

[0082] ;

[0083] This step clears all spectral components outside the BeiDou B3I signal bandwidth to zero, ensuring that subsequent processing energy is concentrated only on the useful signal, and avoiding the problem that a large number of out-of-band noise spectral lines within the 40MHz wide sampling bandwidth are relatively boosted after normalization.

[0084] S5. Perform IFFT (Inverse Frequency-Time Transform):

[0085] right Performing a 1024-point IFFT yields a complex time-domain signal block. .

[0086] S6. Overlapping addition and continuous stream output:

[0087] Because step S2 uses 50% overlap and a Hamming window, the current data block The first 512 points (i=0,...,511) are added to the last 512 points temporarily stored in the previous data block. The result of the addition is the final output, continuous signal after interference suppression. The latest 512 points. Simultaneously, the current data block... The last 512 points (i=512,...,1023) are temporarily stored and used for overlapping and addition with the next data block. This process is repeated to achieve continuous real-time processing.

[0088] Figure 2 is a schematic diagram comparing the effects of narrowband interference suppression before and after using the method of the present invention. Figure 2(a) shows the signal power spectrum before interference suppression. It can be seen that there is a narrowband interference with a bandwidth of 2MHz at the center frequency, whose power spectral density is much higher than that of the navigation signal and noise. Furthermore, the receiver front-end analog filter has a bandwidth of 30MHz; due to the filtering effect, the power spectral density outside ±15MHz is lower than that within ±15MHz. Figure 2(b) shows the power spectrum after interference suppression using the method of the present invention. It can be seen that the originally high interference peaks are significantly "flattened," and the narrowband interference is significantly suppressed. In addition, signals outside the effective frequency band of the navigation signal (-10.23~10.23MHz) are also suppressed.

[0089] After interference suppression according to this application, the signal is acquired, tracked, and the carrier-to-noise ratio (CNR) is estimated using a software receiver. The estimated CNR of the navigation signal is 44.5 dB Hz, a decrease of 1.5 dB compared to the initial CNR. This is because, while suppressing interference, the navigation signal spectral lines overlapping with the interference spectral lines are also suppressed, thus reducing the CNR. Furthermore, if step S4 is not performed (i.e., the out-of-band spectral lines of the BeiDou B3I signal are not zeroed out), the estimated CNR after interference suppression is 40.3 dB Hz, a deterioration of 4.2 dB compared to the method of this invention. This further demonstrates that zeroing out-of-band spectral lines effectively blocks the dilution effect of out-of-band noise, ensuring and optimizing the output signal-to-noise ratio while suppressing interference. Figure 3(a) is the capture correlation peak result obtained after narrowband interference suppression by the method of the present invention. Figure 3(b) is the capture correlation peak result obtained without performing step S4. It can be seen that the signal correlation peak in Figure 3(a) is more obvious than that in Figure 3(b), which is consistent with its higher carrier-to-noise ratio. In the figure, PRN #1 represents pseudo-random noise code 1, amplitude represents amplitude, frequency / Hz represents frequency / hertz, and code phase / samples represents code phase / sampling points.

[0090] It should be understood that although the steps in the flowchart of Figure 1 are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in Figure 1 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0091] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0092] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A real-time method for suppressing narrowband interference in satellite navigation based on power inversion, characterized in that, The method includes: performing segmentation processing on the continuously input satellite navigation digital intermediate frequency signal stream, dividing the signal into consecutive data blocks of a fixed length, applying a window function to each data block to obtain a windowed data block; performing a fast Fourier transform on each windowed data block to convert the time-domain signal to the frequency domain and obtain the corresponding complex spectrum; independently performing a unit circle projection operation on each complex spectrum in the frequency domain to obtain a frequency-domain data block after power inversion; determining the corresponding effective frequency point index range in the digital frequency domain according to the center frequency and nominal bandwidth of the satellite navigation digital intermediate frequency signal, and forcing the spectral lines outside the effective frequency point index range to zero to obtain an optimized frequency-domain data block; performing an inverse fast Fourier transform on the optimized frequency-domain data block to obtain the restored time-domain signal; performing corresponding post-processing on the restored time-domain signal according to the overlap method, recombining and outputting the suppressed time-domain signal; independently performing a unit circle projection operation on each complex spectrum in the frequency domain to obtain a frequency-domain data block after power inversion, including: independently performing a unit circle projection operation on each complex spectrum in the frequency domain, and the obtained frequency-domain data block after power inversion is in, It is a complex spectrum. It is a frequency domain data block after power inversion.

2. The method according to claim 1, characterized in that, The length N of the data block is a power of 2, and the number of overlapping samples L between adjacent data blocks satisfies 0 < L < N. The window function is a Hanning window or a Hamming window, and the window function application formula is: ,in, For the i-th sample in the m-th original data block, Let be the value of the window function at the i-th position. For windowed data blocks.

3. The method according to claim 2, characterized in that, Performing a fast Fourier transform on each windowed data block to convert the time-domain signal to the frequency domain and obtain the corresponding complex spectrum, including: performing a fast Fourier transform on each windowed data block to convert the time-domain signal to the frequency domain, and the obtained corresponding complex spectrum is Where k is the spectral line index. It is a complex spectrum.

4. The method according to claim 1, characterized in that, Determining the corresponding effective frequency point index range in the digital frequency domain according to the center frequency and nominal bandwidth of the satellite navigation digital intermediate frequency signal, including: according to the center frequency of the satellite navigation digital intermediate frequency signal and nominal bandwidth Determine the corresponding effective frequency index range in the digital frequency domain. ,in, Indicates the starting valid frequency point. This indicates the termination of the valid frequency point.

5. The method according to claim 4, characterized in that, The calculation methods of the starting effective frequency point and the ending effective frequency point are respectively in, This is the rounding function. The lower limit frequency of the effective frequency band. Sampling frequency, The upper limit frequency of the effective frequency band. The number of points in the FFT, i.e., the length of the data block.

6. The method according to claim 5, characterized in that, The calculation methods of the lower limit frequency and the upper limit frequency of the effective frequency band are respectively: in, The center frequency of the navigation signal. This is the nominal bandwidth of the navigation signal. Sampling frequency, This is the modulo operation.

7. The method according to claim 1, characterized in that, Forcing the spectral lines outside the effective frequency point index range to zero, including: the process of forcing the spectral lines outside the effective frequency point index range to zero is in, This refers to the frequency domain data after zeroing, i.e., the optimized frequency domain data block. It is a frequency domain data block with inverted power. It is the effective frequency index range of the navigation signal, and k is the spectral line index.

8. The method according to claim 7, characterized in that, Performing an inverse fast Fourier transform on the optimized frequency-domain data block to obtain the restored time-domain signal, including performing an inverse fast Fourier transform on the optimized frequency-domain data block, and the obtained restored time-domain signal is in, This refers to the frequency domain data after zeroing, i.e., the optimized frequency domain data block. The number of points in the FFT. is the index of the time-domain sample, and k is the spectral line index.

9. The method according to claim 1, characterized in that, The segmentation processing adopts the overlap-save method or the overlap-add method.

Images