Data optimization monopulse search method generated based on CCF-ST algorithm

Abstract

The invention relates to the technical field of pulsar search, and discloses a data optimization monopulse search method generated based on a CCF-ST algorithm, and the method comprises the following operation steps: 1, inputting a 19-wave-velocity FITS format file; 2, preprocessing: obtaining a file in an FITS format; step 3, converting the obtained FITS format file into an FIL format file by using Sigproc; 4, judging whether all the FITS format files are converted or not. According to the data optimization monopulse search method generated based on the CCFF-ST algorithm, through accurate marking and screening of interference signals, the number of positive samples is effectively increased, the average error of dispersion is remarkably reduced, and meanwhile, the number of negative samples with the dispersion value larger than 10 is also effectively reduced, so that the quality and efficiency of subsequent data processing are optimized, and the method is suitable for large-scale popularization and application. Particularly, when a positive sample signal with weak local brightness is processed, the interference mark array is adjusted by adopting the CCF-ST algorithm, so that the influence of the interference signal on the type discrimination can be effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of pulsar search, and particularly to a data-optimized single-pulse search method based on the CCF-ST algorithm. Background Art

[0002] Currently, the two main categories of pulsar search methods are single-pulse signal search and periodic signal search. Among them, single-pulse signal search is the mainstream direction. Single-pulse search mainly aims to find strong, non-periodic pulses, which is very suitable for searching for signals that cannot be discovered in periodic signals. The application of single-pulse search technology has led to the important discoveries of rotating radio transients (RRATs) and fast radio bursts (FRBs). In 2006, McLaughlin et al. first discovered RRATs, which are considered a special type of intermittent pulsar. In 2007, Lorimer et al. discovered the first FR in the observational data of the Parkes multibeam pulsar survey (PMPS).

[0003] With the significant improvement in the sensitivity of modern telescopes and the extensive impact of human activities, RFI in radio astronomy has become an important factor that cannot be ignored in single-pulse search. Especially the 500-meter Aperture Spherical Radio Telescope (FAST), as the most sensitive single-aperture radio telescope in the world currently, while enjoying high sensitivity, it is also extremely vulnerable to RFI interference. On the other hand, CRAFTS simultaneously uses multiple digital terminals to collect observational data of multiple scientific targets such as pulsars, neutral hydrogen, molecular spectral lines, transient sources, and FRBs, and generates a huge data volume of approximately 10 PB per year, posing an additional challenge to the data processing ability. It is estimated that the CRAFTS pulsar search will generate tens of thousands to hundreds of thousands of pulsar candidates in each 24-hour sky survey scan. However, through manual diagnosis, it is found that the vast majority of these candidates are actually false signals caused by RFI or cosmic noise, which undoubtedly increases the complexity and difficulty of subsequent analysis work.

[0004] Therefore, for large-scale radio astronomy observation data such as CRAFTS, when using single-pulse search technology to find new pulsars, it is necessary to quickly find an RFI identification method with scientific value and preferentially store its identification results to avoid accumulation, and use a robust single-pulse search candidate identification method to accurately and efficiently distinguish pulsars from RFI. However, currently, HEIMDALL lacks fineness in the preprocessing stage of interference signals, showing a relatively rough processing mode. This limitation further causes the search results of HEIMDALL to be vulnerable to the adverse effects of significant interference signals, thus restricting the accuracy and reliability of its detection results to a certain extent. Summary of the Invention

[0005] (1) Technical Problems to be Solved

[0006] In view of the deficiencies of the prior art, the present invention provides a data-optimized single-pulse search method based on the CCF-ST algorithm, which has the advantages of an identification accuracy rate increase of 2.85%, and solves the problem that currently HEIMDALL lacks fineness in the preprocessing stage of interference signals, showing a relatively rough processing mode. This limitation further causes the search results of HEIMDALL to be vulnerable to the adverse effects of significant interference signals, thus restricting the accuracy and reliability of its detection results to a certain extent.

[0007] (2) Technical Solutions

[0008] To achieve the above object of an identification accuracy rate increase of 2.85%, the present invention provides the following technical solutions: A data-optimized single-pulse search method based on the CCF-ST algorithm, including the following operating steps:

[0009] Step 1: Input a FITS format file with a wave speed of 19;

[0010] Step 2: Preprocessing: Obtain a FITS format file;

[0011] Step 3: Use Sigproc to convert the obtained FITS format file into a FIL format;

[0012] Step 4: Whether all FITS format files have been converted;

[0013] Step 5: Input 19 converted FIL files into the CCF-ST algorithm;

[0014] Step 6: Generate 19 mask.dat interference marker files;

[0015] Step 7: Modify the interference marker group number of Heimdall according to the interference marker file;

[0016] Step 8: Single-pulse signal search;

[0017] Step 9: Generate cand files and corresponding H5 pictures;

[0018] Step 10: Manual review and mark the results.

[0019] Preferably, the FITS format file described in Step 1 is obtained by the CRAFTS19 beam receiver. In the CRAFTS19 beam FITS file data, it is mainly affected by the following three types of RFI:

[0020] The first type is narrowband RFI: This type of interference source is widespread and may come from various sources, and may even originate from the instrument itself.

[0021] The second type is 1MHz-wide RFI: This RFI is caused by standing waves and usually shows an irregular time-frequency distribution pattern in the L band, with strong local characteristics.

[0022] The third type is fixed-frequency RFI: It usually comes from satellites or civil aviation aircraft and appears as strong interference at a fixed frequency, with a wide time-frequency distribution.

[0023] Preferably, the FIL file format described in Step 3 is converted from the original FITS file by SIGPROC software, and the FIL file format is much simpler in structure compared to the FITS file format.

[0024] Preferably, the frequency-time data in the FIL file described in Step 3 is saved in the form of a binary stream, and the data reading process can be implemented through file streams in the C++ programming language.

[0025] Preferably, when reading a 2bits FIL file in the C++ programming, it is necessary to calculate a predetermined step size and read data according to the step size.

[0026] Preferably, the mask.da interference marking file described in Step 6 is processed by HEIMDALL software. In HEIMDALL software, the source file corresponding to the interference removal module is clean_filterbank_rfi.cu. The parameter used to mark the interference channels is an array of the C language type. The part that actually calls this parameter is located in the Pipeline.cu module, and the vector class template in the C++ Standard Template Library is used, with the variable name h_killmask. The length of this marking array is the same as the number of frequency channels in the input FIL file.

[0027] Preferably, the specific operation in Step Nine is as follows: Use a Python script to convert the.cand files generated by all beam files with dispersion values close to the true signal dispersion value into.csv format, and further generate an H5 file through YOUR script. Subsequently, use the your_h5ploter.py script to manually review the PNG images corresponding to the generated H5 files.

[0028] (III) Beneficial Effects

[0029] Compared with the prior art, the present invention provides a data-optimized monopulse search method based on the CCF-ST algorithm, which has the following beneficial effects:

[0030] 1. For the data-optimized monopulse search method based on the CCF-ST algorithm, the adjustment of the interference marking array plays an important role in the identification and suppression of RFI signals. Through the precise marking and screening of interference signals, the number of positive samples is effectively increased, the average error of dispersion is significantly reduced, and at the same time, the number of negative samples with a dispersion value greater than 10 is also effectively reduced, thereby optimizing the quality and efficiency of subsequent data processing. Especially when processing positive sample signals with relatively weak local brightness, using the CCF-ST algorithm to adjust the interference marking array can effectively reduce the influence of interference signals on their type discrimination. However, when processing positive sample signals with strong local interference characteristics, the adjustment of the interference marking array fails to effectively identify the positive and negative properties of the signals, indicating that when interference is mixed with positive sample signals, CCF-ST may misjudge positive samples as interference signals. Based on a reasonable interference marking and removal strategy, the integrity of the true signal can be guaranteed to the greatest extent, and the influence of RFI can be effectively suppressed, thereby improving the accuracy and reliability of the overall signal processing process.

[0031] 2. For the data-optimized monopulse search method based on the CCF-ST algorithm, by combining the CCF-ST algorithm with the HEIMDALL software, in the application of multi-beam data in the data-optimized monopulse search process using the CCF-ST algorithm, this method significantly improves the recognition accuracy of positive samples, with an increase of 2.85%. At the same time, on the M06 beam and M19 beam containing known source signals, this method effectively reduces the misjudgment rate of negative samples, by 13.4% and 5.56% respectively, thereby further enhancing the performance of interference identification and pulsar detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the basic process of the experimental treatment of the present invention;

[0033] Figure 2 It is a schematic diagram of the CCF-ST algorithm process of the present invention;

[0034] Figure 3 Schematic diagram of the CRAFTS19 beam position map information of the present invention;

[0035] Figure 4 Schematic diagram of the CFITSIO data reading process of the present invention;

[0036] Figure 5 Schematic diagram of the 2bitsFIL data reading process by the C++ program of the present invention;

[0037] Figure 6 Flow chart of the HEIMDALL processing of the present invention;

[0038] Figure 7 Schematic diagram of the result of modifying the Heimdall interference marker array of the present invention;

[0039] Figure 8 Schematic diagram of the result of not modifying the Heimdall interference marker array of the present invention;

[0040] Figure 9 Known source map with weak signal brightness of the present invention;

[0041] Figure 10 Schematic diagram of the known source signal of the present invention. Specific implementation mode

[0042] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work belong to the scope of protection of the present invention.

[0043] Please refer to Figure 1-10 , a data optimization monopulse search method based on the CCF-ST algorithm, including the following operation steps:

[0044] Step 1: Input a FITS format file with 19 wave speeds;

[0045] Step 2: Pretreatment: Obtain a FITS format file;

[0046] Step 3: Use Sigproc to convert the obtained FITS format file into a FIL format;

[0047] Step 4: Whether all FITS format files are converted;

[0048] Step 5: Input 19 converted FIL files into the CCF-ST algorithm;

[0049] Step 6: Generate 19 mask.dat interference marker files;

[0050] Step 7: Modify the number of interference marker groups in Heimdall according to the interference marker files;

[0051] Step 8: Single-pulse signal search;

[0052] Step 9: Produce cand files and corresponding H5 pictures;

[0053] Step 10: Manually review and mark the results.

[0054] In the case implementation, in Step 1, the FITS format adopts the CCF-ST algorithm. The CCF-ST algorithm is an RFI identification algorithm designed specifically for data in the multi-beam FITS file format. The FITS format files are obtained by the CRAFTS19 beam receiver. The basic characteristics of the FITS file format data are shown in Table 1. Taking 2-bit data as an example, each data file contains 128 sub-integrations, each sub-integration contains 256 spectra, and each spectrum has 4096 frequency channels, thus forming a time-frequency image with a size of (4096×32768). The data can be read through the API interface provided by the CFITSIO library;

[0055] Table 1 The basic information of the FITS file is as follows:

[0056]

[0057] Among them, in the CRAFTS19 beam FITS file data, it is mainly affected by the following three types of RFI:

[0058] The first is narrowband RFI: This type of interference source is widespread and may come from various sources, and may even originate from the instrument itself. However, since the installation of electromagnetic shielding in 2019, the impact of narrowband RFI has been significantly reduced.

[0059] The second is 1MHz-wide RFI: This type of RFI is caused by standing waves and usually shows an irregular time-frequency distribution pattern in the L band, with strong local characteristics.

[0060] The third is fixed-frequency RFI: It usually originates from satellites or civil aviation aircraft and appears as strong interference at fixed frequencies, with a wide time-frequency distribution.

[0061] In addition, the CRAFTS data may also have short-time, narrow-bandwidth types of RFI from unknown sources, whose interference duration is short, the bandwidth is small, and it usually has high local characteristics;

[0062] In the case implementation, the FIL file format is converted from the original FITS file through the SIGPROC software, and the FIL file format is much simpler in structure than the FITS file format. Its structure is shown in Table 3. It should be noted that in the psrfits2fb.c source code of the SIGPROC software, there are specific requirements for the number of polarizations of the original FITS file. Specifically, when the number of polarizations of the original FITS file exceeds single polarization, direct conversion operations will no longer be supported. In this case, in order to meet the conversion requirements, the number of polarizations of the original FITS file must be merged to ensure that its number of polarizations meets the limitations of the conversion process;

[0063] The basic format of the FIL file is as follows in Table 3:

[0064]

[0065] In the case implementation, the basic data information of the FIL file is shown in Table 4. By comparing with the FITS information in Table 1, it can be seen that in the process of converting to the FIL file format, the SIGPROC software does not use the SUBINT (sub-integration) information of the FITS file;

[0066] The basic information of the FIL file is as follows in Table 4:

[0067]

[0068] In the case implementation, in the pulsar search software PRESTO, the SUBINT information in the FIL file can be read through the readfile program, as shown in Table 2. However, the SUBINT information extracted by readfile is not consistent with the information in the original FITS file. Specifically, when the readfile program reads files in different formats, due to the differences in the information contained in different file formats, the SUBINT information of the FIL file it returns is not real data, but a fixed value designed to be compatible with the reading requirements of different formats. This design is mainly to simplify the writing and implementation of the program so that it can adapt to the reading of multiple file formats. Table 2 shows some examples of the program's reading results;

[0069] The partial information of readfile for the FITS file and the FIL file is as follows in Table 2:

[0070]

[0071] In the case implementation, the basic principle of the CCF-ST algorithm is based on the different characteristics of RFI and pulsar signals. Specifically, RFI can be regarded as a near-field extended source, and its signals can be received in multiple beam directions, while pulsar signals are typical point sources, and their signals can only be received in specific directions. Considering that the CRAFTS19 beam data receives signals from multiple beam positions simultaneously, RFI and pulsar signals can be effectively distinguished by analyzing the similarity between the CRAFTS19 beam signals. The position distribution of the 19 beams is as Figure 3 shown, and the working process of the CCF-ST algorithm is as Figure 2 shown, which is mainly divided into three steps: First, calculate the correlation coefficient of the signals through the cross-correlation function (CCF) between each beam; Second, evaluate the quantization value of RFI according to the calculated correlation coefficient; Finally, use the SumThreshold algorithm to mark and screen RFI;

[0072] Among them, the input module of the CCF-ST algorithm is currently limited to receiving FITS format files, while the Heimdall software processing process uses FIL format files, and there is an inconsistency in the data storage order of these two format files. Therefore, in order to make the CCF-ST algorithm compatible with and process FIL format files, we need to reconstruct the input processing module of the CCF-ST algorithm. This reconstruction process aims to ensure that the algorithm can accurately parse the specific structure of the FIL format file, so as to effectively extract and utilize the data therein.

[0073] In the case implementation, based on the basic structure of the FIL file shown in Table 3, we can reasonably infer that the frequency-time data in the FIL file is saved in the form of a binary stream. In view of this characteristic, the data reading process can be implemented through the file stream in the C++ programming language. This is specifically reflected and applied in the source code of the HEIMDALL software package, especially in the SigProcFile.cpp source file. The SIGPROC official documentation points out that in the data conversion process, the SIGPROC program adopts a specific order for data access and storage, so there is a significant difference in the reading order from the API interface provided by the CFITSIO library when using the file stream. Taking the reading of 2-bit data as an example, Figure 4 shows the specific process of the CFITSIO library API interface reading the frequency-time matrix;

[0074] Among them, the API interface provided by the CFITSIO library allows the caller to control how much data is read at one time. From Figure 4As can be seen, the order in which the API interface provided by CFITSIO reads data is to read data from the first time sampling point into the array in sequence from low frequency to high frequency in the frequency dimension. The process of reading data from a 2-bit FIL file by the C++ file stream is as Figure 5 shown;

[0075] When reading a 2-bit FIL file in a C++ program, it is necessary to calculate a predetermined step size and read data according to the step size. Specifically, the program first calculates the step size and starts reading data from the high-frequency channel determined by the corresponding step size. This process results in the order of the read data being inconsistent with the order of the data in the original FITS file. Therefore, in order to restore the original order of the data, it is necessary to perform a reordering process on the displayed read data to make it consistent with the data order in the FITS file.

[0076] In the case implementation, the basic process of processing by the HEIMDALL software is as Figure 6 shown, and the main modification to it is the part of marking the mask in the interference cancellation module;

[0077] Among them, in the HEIMDALL software, the source file corresponding to the interference cancellation module is clean_filterbank_rfi.cu. The parameter used to mark the interference channels is an array of the C language type. The part that actually calls this parameter is located in the Pipeline.cu module and uses the vector class template in the C++ Standard Template Library (STL), with the variable name h_killmask. The length of this marking array is the same as the number of frequency channels in the input FIL file. However, the data volume of the mask.dat file is related to the product of the number of SUBINTs and the number of frequency channels;

[0078] Therefore, the data in the mask.dat file is read into a two-dimensional array, where the rows of the array represent the number of SUBINTs, the columns represent the number of frequency channels, and the frequencies are arranged from low to high. To determine whether each frequency channel is interfered, this paper accumulates each column in row-major order. When the accumulated value of a column exceeds half of the number of SUBINTs, it indicates that the frequency channel is interfered, and this channel is marked as 0 in the marking array (in the HEIMDALL software, the marking value of the interfered channel is 0, and that of the non-interfered channel is 1). Finally, the HEIMDALL software records the frequency information in the FIL file by specifying the highest frequency value and the negative difference between frequencies for recursion. Therefore, the HEIMDALL software processes the data from high frequency to low frequency. Thus, it is necessary to reverse the marked mask array to match the processing order of the HEIMDALL software, and this process can be achieved through the reverse() function in C++.

[0079] In the case implementation, when analyzing the number of candidate bodies generated with or without modifying the HEIMDALL marking array, the FIL file of the CRAFTS19 beam containing the known source signal is input into the CCF-ST algorithm, and it is verified through a control experiment whether to modify the interference marking array in HEIMDALL. The generated.mask.dat file and the corresponding.cand file are compared. The experimental results are shown in Table 5. The comparative analysis shows that after modifying the interference marking array, the number of candidate bodies generated by HEIMDALL has changed significantly;

[0080] Table 5 Comparison of the number of candidate bodies generated is as follows:

[0081]

[0082]

[0083] In the case implementation, by using a Python script, the cand files generated from all beam files with dispersion values close to the true signal are converted into csv format, and then an H5 file is further generated through YOUR script. Subsequently, the your_h5ploter.py script is used to manually review the PNG graph corresponding to the generated H5 file. The analysis results show that after updating the interference marking array using the mask.dat file, the number of positive samples has increased by 2.85% compared to when not updated. The known source signal and its changes are shown in Table 6;

[0084] Table 6 Comparison of the number of positive samples generated is as follows:

[0085]

[0086]

[0087] In the case implementation, the dispersion values of all positive sample signals were statistically analyzed, and the mean absolute error between them and the dispersion values provided by ATNF was calculated. The results are shown in Table 7. The analysis results indicate that after modifying the interference marking array, the error of the dispersion value was significantly reduced;

[0088] The comparison of the average values of the positive sample dispersion errors is as follows in Table 7:

[0089]

[0090] In the case implementation, the quantity information of all negative samples whose dispersion values generated from the beam files containing positive samples were greater than 10 was statistically analyzed. The results are shown in Table 8. The analysis results indicate that after modifying the interference marking array of Heimdall, the number of negative samples was significantly reduced compared with that without modifying the interference marking array of Heimdall;

[0091] The comparison of the number of negative samples is as follows:

[0092]

[0093]

[0094] In the case implementation, whether the interference marking array was modified or not did not have a significant impact on the recognition of known sources with strong signal brightness. This result indicates that regardless of whether the interference marking array was adjusted or not, the known source map with strong signal brightness could be accurately recognized. The results of modifying the Heimdall interference marking array are as Figure 7 shown, and the results without modifying the Heimdall interference marking array are as Figure 8 shown.

[0095] In the case implementation, for some known source maps with weak signal brightness generated by beams, after modifying the interference marking array of Heimdall, the relevant signal maps could be successfully recognized and drawn, while when the interference marking array was not modified, the relevant signal maps could not be successfully recognized and drawn. This result indicates that the adjustment of the interference marking array has a better effect in the recognition of known sources with weak signal brightness. Figure 9 Such known source maps with weak signal brightness are shown.

[0096] In the case implementation, for some known source signals generated by beams but with strong narrowband interference beside them, they were successfully recognized and drawn after not modifying the interference marking array of Heimdall, while they could not be successfully recognized after modifying the interference marking array of Heimdall. This result indicates that if there is strong interference near the known source, modifying the marking array does not achieve good results. This type of known source signal is asFigure 10 as shown

[0097] In summary, for the data optimization monopulse search method based on the CCF-ST algorithm, the adjustment of the interference marking array plays an important role in the recognition and suppression of RFI signals. Through the precise marking and screening of interference signals, the number of positive samples is effectively increased, the average error of dispersion is significantly reduced, and at the same time, the number of negative samples with a dispersion value greater than 10 is also effectively reduced, thereby optimizing the quality and efficiency of subsequent data processing. Especially when processing positive sample signals with relatively weak local brightness, adjusting the interference marking array using the CCF-ST algorithm can effectively reduce the influence of interference signals on their type discrimination. However, when processing positive sample signals with strong local interference characteristics, the adjustment of the interference marking array fails to effectively identify the positive and negative nature of the signals, indicating that when interference mixes with positive sample signals, CCF-ST may misjudge positive samples as interference signals. Based on a reasonable interference marking and removal strategy, the integrity of real signals can be guaranteed to the greatest extent, and the influence of RFI can be effectively suppressed, thereby improving the accuracy and reliability of the overall signal processing process.

[0098] Moreover, by combining the CCF-ST algorithm with the HEIMDALL software and using the data optimization monopulse search process generated by the CCF-ST algorithm, in the application of multi-beam data, this method significantly improves the recognition accuracy of positive samples, with an increase of 2.85%. At the same time, on the M06 beam and M19 beam containing known source signals, this method effectively reduces the misjudgment rate of negative samples, by 13.4% and 5.56% respectively, thereby further enhancing the performance of interference recognition and pulsar detection, and solving the problem that HEIMDALL currently lacks fineness in the preprocessing stage of interference signals and shows a relatively rough processing mode. This limitation further causes the search results of HEIMDALL to be vulnerable to the adverse effects of significant interference signals, thereby restricting the accuracy and reliability of its detection results to a certain extent.

[0099] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the element.

[0100] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A data-optimized monopulse search method based on the CCF-ST algorithm, characterized in that: It includes the following operation steps: Step 1: Input the FITS format file with a wave speed of 19; Step 2: Preprocessing: Obtain a FITS format file; Step 3: Use Sigproc to convert the obtained FITS format file into FIL format; Step 4: Whether all FITS format files are converted; Step 5: Input the 19 converted FIL files into the CCF-ST algorithm; Step 6: Generate 19 mask.dat interference marker files; Step 7: Modify the interference marker group number of Heimdall according to the interference marker file; Step 8: Single-pulse signal search; Step 9: Generate cand files and corresponding H5 pictures; Step 10: Manually review and mark the results.

2. The data optimization monopulse search method based on the CCF-ST algorithm according to claim 1, characterized in that: The FITS format file described in Step 1 is obtained by the CRAFTS19 beam receiver. In the data of the CRAFTS19 beam FITS file, it is mainly affected by the following three types of RFI: The first is narrowband RFI: This type of interference source is widespread and may come from various sources, and may even originate from the instrument itself. The second is 1MHz-wide RFI: This RFI is caused by standing waves and usually shows an irregular time-frequency distribution pattern in the L band, with strong local characteristics. The third is fixed-frequency RFI: It usually originates from satellites or civil aviation aircraft and appears as strong interference at a fixed frequency, with a wide time-frequency distribution.

3. The data-optimized monopulse search method based on the CCF-ST algorithm according to claim 1, characterized in that: The FIL file format described in Step 3 is converted from the original FITS file by the SIGPROC software, and the FIL file format is much simpler in structure than the FITS file format.

4. A data-optimized monopulse search method based on the CCF-ST algorithm according to claim 1, characterized in that: The frequency-time data in the FIL file described in Step 3 is saved in the form of a binary stream, and the data reading process can be implemented through the file stream in the C++ programming language.

5. A data optimization monopulse search method based on the CCF-ST algorithm according to claim 4, characterized in that: When the C++ programming reads the 2bits FIL file, it is necessary to calculate a predetermined step size and read the data according to the step size.

6. A data-optimized monopulse search method based on the CCF-ST algorithm according to claim 1, characterized in that: The mask.da interference marker file described in Step 6 is processed by the HEIMDALL software. In the HEIMDALL software, the source file corresponding to the interference removal module is clean_filterbank_rfi.cu. The parameter used to mark the interference channel is an array of the C language type. The part that actually calls this parameter is located in the Pipeline.cu module and uses the vector class template in the C++ standard template library, with the variable name h_killmask. The length of this marker array is the same as the number of frequency channels in the input FIL file.

7. A data optimization monopulse search method based on the CCF-ST algorithm according to claim 1, characterized in that: The specific operation described in Step 9 is as follows: Use a Python script to convert the.cand files generated by all beam files with a dispersion value similar to the real signal into the.csv format, and further generate H5 files through the YOUR script. Subsequently, use the your_h5ploter.py script to manually review the PNG images corresponding to the generated H5 files.

Images