A method for generating ultra-wideband pulsar baseband data

Abstract

The present invention provides a method for generating ultra-wideband pulsar baseband data, comprising: reading in original pulsar baseband data in blocks according to time; performing FFT on each data block to obtain frequency domain data; performing coherent de-dispersion operation on the frequency domain data using the bandwidth, center frequency and dispersion of the original pulsar baseband data to obtain the Fourier transform result of the signal propagated by the pulsar; setting the target bandwidth, target center frequency and target dispersion according to requirements, adding dispersion through the inverse operation of coherent de-dispersion, obtaining the Fourier transform result of the new pulsar signal voltage, and performing IFFT to obtain ultra-wideband pulsar baseband data. The method of the present invention is based on pulsar baseband data with a narrow bandwidth, performs a coherent de-dispersion operation to remove the dispersion of the template baseband data and uses its inverse operation to add dispersion, so as to ensure that the generated ultra-wideband pulsar baseband data has complete characteristics, so that the generated ultra-wideband pulsar baseband data has more research significance.

Description

Technical Field

[0001] The invention belongs to the field of radio astronomy, and in particular relates to a pulsar baseband data generation method, which is specially used for generating ultra-wideband pulsar signals. Background Art

[0002] With the rapid development of manufacturing technology and information technology, radio telescopes have developed in the direction of multi-array, large aperture and ultra-wideband. Array and ultra-wideband receiving related equipment have dramatically increased the amount of data collected. For example, all antennas of the Low-Frequency Array (LOFAR) generate 14Tb / s of raw data per second at a sampling rate of 200MHz. The data sampling rate of the Five-hundred-meter Aperture Spherical Telescope (FAST) terminal reaches 12.8GB / s with a time resolution of 100 microseconds, 4000 channels, dual polarization and 19 beams. The annual archived data generated by the ultra-wideband and PAF receiving systems of the 110-meter omnidirectional movable radio telescope (Qi Tai Radio Telescope, QTT) under construction in Qitai, Xinjiang is expected to exceed 10PB.

[0003] With the continuous deepening of astronomical research and the rapid development of digital technology, the performance requirements of astronomical observation research on digital terminal system-related equipment are also constantly increasing. The terminal system needs to achieve high-speed sampling, real-time analysis and data preprocessing at a wider bandwidth, higher time resolution and higher frequency resolution. The use of PAF and multi-beam receiving systems has doubled the amount of information obtained from astronomical observations. The transmission and storage of the generated high-speed data streams and how to achieve real-time processing of massive astronomical data on heterogeneous platforms are currently urgent problems that need to be solved during the operation of various radio observation equipment.

[0004] The demand for wider sampling bandwidth, higher digital signal bit width and more array antennas has led to an exponential growth in the amount of data to be processed. Due to the performance limitations of storage devices, massive astronomical signals can only be processed and analyzed in real time, which poses a huge challenge to computing hardware and software.

[0005] Currently, mainstream digital terminal systems mostly use a hybrid architecture of FPGA, CPU and GPU. Heterogeneous systems require data to be exchanged between different devices. One of the technical difficulties of data processing systems based on heterogeneous platforms is how to achieve high-speed data flow between the CPU and GPU and avoid problems such as high bit error rate and high packet loss rate during massive data transmission. For high-speed data transmission and distribution, a feasible method is to open a ring buffer in the memory, cache the data in real time, and transfer the cached data to the GPU memory, ultimately realizing real-time data processing in the GPU.

[0006] The use of graphics processors can significantly improve the running speed of applications in scientific analysis, simulation, etc. Internationally, the use of CPU+GPU heterogeneous computing platforms for real-time processing of pulse signals has become the mainstream. Using the computing unit of the GPU to process observation data can reduce the CPU load and greatly enhance the data flow processing efficiency of the system. With the continuous development of radio astronomy digital terminal technology, the real-time processing of signals has posed a severe challenge to traditional computing technology. The powerful computing power of GPU clusters provides a feasible solution for processing the massive data streams generated during radio astronomy observations.

[0007] In the process of astronomical technology research, ultra-wideband receivers and related digital signal processing algorithms are often not carried out synchronously. The premise of collecting ultra-wideband pulsar baseband data is to have an ultra-wideband receiver. However, at present, only a few radio telescopes at home and abroad are equipped with ultra-wideband receivers. The lack of ultra-wideband pulsar baseband data is not conducive to the early algorithm verification. Specifically, for astronomical observatories without ultra-wideband receivers, researchers cannot effectively carry out research on ultra-wideband pulsar baseband data algorithms. Therefore, a new method is needed to simulate and generate ultra-wideband pulsar baseband data. Summary of the invention

[0008] The object of the present invention is to provide a method for generating ultra-wideband pulsar baseband data to ensure that the generated ultra-wideband pulsar baseband data has complete characteristics, thereby making the generated ultra-wideband pulsar baseband data more meaningful for research.

[0009] In order to achieve the above object, the present invention provides a method for generating ultra-wideband pulsar baseband data, comprising:

[0010] S1: Read the original pulsar baseband data in blocks according to time to obtain data blocks;

[0011] S2: Perform FFT on each data block to obtain frequency domain data as the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the time domain;

[0012] S3: Using the bandwidth, center frequency and dispersion of the original pulsar baseband data, a coherent de-dispersion operation is performed on the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the time domain to obtain the Fourier transform result of the signal propagated by the pulsar in the time domain;

[0013] S4: Set the target bandwidth, target center frequency and target dispersion amount according to the requirements, add dispersion through the inverse operation of the coherent dedispersion operation according to the target bandwidth, target center frequency and target dispersion amount, obtain the Fourier change result of the pulsar signal voltage obtained by the radio telescope observation in the new time domain, and perform IFFT to obtain ultra-wideband pulsar baseband data.

[0014] In step S3, a coherent de-dispersion operation is performed on the Fourier change result of the pulsar signal voltage obtained by radio telescope observation in the time domain using the bandwidth, center frequency and dispersion of the original pulsar baseband data, specifically including:

[0015] S31: Obtain the inverse of the transfer function using the bandwidth, center frequency and dispersion of the original pulsar baseband data;

[0016] S32: Obtaining a Fourier transform result of a signal propagated by the pulsar in the time domain according to the inverse of the transfer function and a Fourier transform result of a pulsar signal voltage obtained by radio telescope observation in the time domain;

[0017] In step S4, dispersion is added by an inverse operation of the coherent de-dispersion operation according to the target bandwidth, the target center frequency and the target dispersion amount, and the Fourier change result of the pulsar signal voltage obtained by the radio telescope observation in the new time domain is obtained, which specifically includes:

[0018] S41: obtaining a target transfer function according to a target bandwidth, a target center frequency and a target dispersion amount;

[0019] S42: Based on the target transfer function and the Fourier transform result of the signal propagated by the pulsar in the time domain V int (f), and obtain the Fourier transform result V′(f) of the pulsar signal voltage obtained by radio telescope observation in the new time domain.

[0020] The inverse of the transfer function H(f freq +Δf) -1 for:

[0021]

[0022] Among them, f ref is the center frequency, Δf is the difference between the observed frequency f and the center frequency f ref The distance; f ref+Δf is the observation frequency f, and its variation range is from the minimum frequency to the maximum frequency of the band, and the size of this variation range is the bandwidth; D is the dispersion constant, and DM is the dispersion amount;

[0023] The signal v propagated by a pulsar in the time domain int The Fourier transform result V of (t) int (f) is:

[0024] V int (f)=V(f)H(f) -1 ,

[0025] Among them, V(f), V int (f) and H(f) are v(t), v int (t) and the Fourier transform result of h(t), H(f) -1 is the inverse of the transfer function, and f is the observation frequency.

[0026] The target transfer function H′(f freq +Δf) is:

[0027]

[0028] The Fourier transform result V′(f) of the pulsar signal voltage obtained by radio telescope observation in the new time domain is:

[0029] V′(f)=V int (f)H′(f),

[0030] Where H′(f) is the target transfer function, V int (f) is the Fourier transform result of the signal propagated by the pulsar in the time domain, and V′(f) is the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the new time domain.

[0031] In the step S1, the original pulsar baseband data is original pulsar baseband data in psrdada format and is time domain data;

[0032] In step S3, the values ​​of the bandwidth, center frequency and dispersion amount of the original pulsar baseband data are specifically obtained by reading the data header part of the original pulsar baseband data in the psrdada format.

[0033] The method for generating ultra-wideband pulsar baseband data also includes step S5: packaging the ultra-wideband pulsar baseband data into a psrdada format.

[0034] When packaging into the psrdada format, the center frequency of the data header part of the psrdada format is changed to the target center frequency, the bandwidth is changed to the target bandwidth, the dispersion amount is changed to the target dispersion amount, and the sampling interval is changed to 1 / (2×target bandwidth) milliseconds.

[0035] The step S5 further includes: calculating the folding period of the ultra-wideband pulsar baseband data finally generated according to the folding period of the original pulsar baseband data, the bandwidth of the folding period of the original pulsar baseband data and the target bandwidth.

[0036] The bandwidth of the original pulsar baseband data is above 100 MHz, and the target bandwidth is above 1 GHz.

[0037] The coherent de-dispersion operation and the inverse operation of the coherent de-dispersion operation are implemented using a GPU.

[0038] The method of the present invention is based on pulsar baseband data with a narrow bandwidth, performs a coherent de-dispersion operation to remove the dispersion of template baseband data, and uses its inverse operation to add dispersion, thereby ensuring that the generated ultra-wideband pulsar baseband data has complete characteristics and can retain information such as pulsar dispersion, so that the generated ultra-wideband pulsar baseband data has more research significance, so that it can be used to test the correctness of radio astronomy terminals and related algorithms, and improve the work efficiency of pulsar baseband processing algorithm research, which cannot be achieved by other simulation methods.

[0039] In addition, the present invention packages the ultra-wideband pulsar baseband data into a psrdada format, and the generated ultra-wideband pulsar baseband data can be processed by professional pulsar processing software.

[0040] The present invention sets the target bandwidth, target center frequency and target dispersion according to the requirements, so it can generate customized ultra-wideband pulsar baseband data according to its own needs.

[0041] The algorithm data processing speed has a lot of room for improvement and is easy to implement on GPU. The de-dispersion and addition parts are implemented on GPU, which is easy to parallelize and thus improve the processing speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flow chart of the method for generating ultra-wideband pulsar baseband data of the present invention.

[0043] Figure 2A-2B It is the phase spectrum and pulse profile of 400MHz broadband baseband data before de-dispersion.

[0044] Figure 3A-3B It is the phase spectrum diagram and pulse contour diagram of 400MHz broadband baseband data after dispersion elimination.

[0045] Figure 4A-4B It is the phase spectrum diagram and pulse contour diagram of 3328MHz broadband baseband data before dispersion elimination.

[0046] Figure 5A-Figure 5B It is the phase spectrum diagram and pulse contour diagram of 3328MHz broadband baseband data after de-dispersion. DETAILED DESCRIPTION

[0047] An embodiment of the present invention is given below in conjunction with the accompanying drawings to further explain and illustrate the scheme of the present invention in detail.

[0048] like Figure 1 The figure shows a flow chart of the method for generating ultra-wideband pulsar baseband data. Figure 1 As shown, the method for generating ultra-wideband pulsar baseband data includes:

[0049] Step S1: Read the original pulsar baseband data in blocks according to time to obtain data blocks;

[0050] The original pulsar baseband data is in psrdada format and is time domain data. Thus, data generation is performed using the original pulsar baseband data with a narrow bandwidth as a template.

[0051] Step S2: Perform FFT on each data block to obtain frequency domain data (i.e., convert the time domain to the frequency domain) to facilitate subsequent coherent de-dispersion, as the Fourier transform result V(f) of the pulsar signal voltage obtained by radio telescope observation in the time domain;

[0052] Step S3: Using the bandwidth bw, center frequency cfreq and dispersion dm of the original pulsar baseband data, a coherent de-dispersion operation is performed on the Fourier transform result V(f) of the pulsar signal voltage obtained by radio telescope observation in the time domain to obtain the Fourier transform result V of the signal propagated by the pulsar in the time domain. int (f).

[0053] In this embodiment, the bandwidth, center frequency and dispersion value of the original pulsar baseband data are specifically obtained by reading the data header part of the original pulsar baseband data in the psrdada format.

[0054] The dispersion effect of the interstellar medium on radio waves is equivalent to the radio waves passing through a filter with the interstellar medium transmission function. Coherent dedispersion can eliminate the dispersion effect caused by the interstellar medium (ISM) in the signal by passing the observed signal through a filter with the inverse of the interstellar medium (ISM) transmission function.

[0055] Suppose the signal transmitted by the pulsar in the time domain is v int(t), the pulsar signal voltage obtained by radio telescope observation is v(t), and the response function of the interstellar medium is h(t), then the pulsar signal voltage v(t) obtained by radio telescope observation is:

[0056] v(t)=v int (t)*h(t) (1)

[0057] Based on the convolution theorem, the convolution of a sequence in the time domain is equivalent to the product of a sequence in the frequency domain, that is,

[0058] V(f)=V int (f)H(f) (2)

[0059] V(f), V int (f) and H(f) are the pulsar signal voltage v(t) obtained by radio telescope observation in the time domain and the pulsar propagation signal v in the time domain, respectively. int (t) and the Fourier transform result of the interstellar medium response function h(t), where H(f) is the transfer function and f is the observation frequency.

[0060] Set the center frequency f ref , a certain observation frequency f and the center frequency f ref The distance Δf is:

[0061] Δf=ff ref (3)

[0062] That is, formula (2) can be written as:

[0063] V(f ref +Δf)=V int (f ref +Δf)H(f ref +Δf) (4)

[0064] The phase φ(f) acquired by an electromagnetic wave propagating in the interstellar medium is:

[0065]

[0066] Where L is the distance of the ionized plasma from the Earth, f p is the plasma frequency of the interstellar medium, c is the speed of light, and f is the observation frequency, based on the center frequency f ref The transfer function H(f ref +Δf) can be written as:

[0067] H(f ref +Δf)=e -iφ(fref+Δf) (6)

[0068] Using Taylor expansion, based on the center frequency f refThe phase φ(f ref +Δf) (i.e., the phase φ(f) acquired by the electromagnetic wave propagating in the interstellar medium) can be written as:

[0069]

[0070] The first term in the square brackets in equation (7) is a constant phase offset that is lost in the detection, while the second term is a linear frequency gradient, equivalent to the time delay of the pulse arrival, and the third term causes in-band dispersion that needs to be corrected. The transfer function to correct in-band dispersion is then the inverse of the third term in equation (7).

[0071] For the dispersion measurement of pulsars, based on the center frequency f ref The transfer function H(f ref +Δf) only retains the center frequency f ref The phase φ(f ref +Δf), it can be expressed as:

[0072]

[0073] Accordingly, formula (8) can be rewritten as:

[0074]

[0075] Among them, f ref is the center frequency, Δf is the difference between the observed frequency f and the center frequency f ref The distance, f ref The variation range of +Δf is from the minimum frequency to the maximum frequency of the frequency band, and the size of this variation range is the bandwidth, D is the dispersion constant, and DM is the dispersion amount.

[0076] Among them, D (dispersion constant) and DM (dispersion amount) are known parameters.

[0077]

[0078]

[0079] Where e represents the elementary charge, m e represents the mass of the electron, c represents the speed of light, and d represents the distance between the pulsar and the earth; n e is the electron density in interstellar space, and l is the distance from the observed pulsar to the radio telescope.

[0080] The inverse of the transfer function H(f freq +Δf) -1 for:

[0081]

[0082] Among them, f ref is the center frequency, Δf is the difference between the observed frequency f and the center frequency f ref The distance; f ref +Δf is the observation frequency f, and its variation range is from the minimum frequency to the maximum frequency of the frequency band, and the size of this variation range is the bandwidth; D is the dispersion constant, and DM is the dispersion amount.

[0083] Therefore, the bandwidth, center frequency and dispersion of the original pulsar baseband data are used to perform coherent de-dispersion operations on the Fourier change results of the pulsar signal voltage obtained by radio telescope observations in the time domain, including:

[0084] Step S31: Obtain the inverse of the transfer function using the bandwidth, center frequency and dispersion of the original pulsar baseband data;

[0085] As mentioned above, the inverse of the transfer function H(f freq +Δf) -1 for:

[0086]

[0087] Among them, f ref is the center frequency, Δf is the difference between the observed frequency f and the center frequency f ref The distance; f ref +Δf is the observation frequency f, and its variation range is from the minimum frequency to the maximum frequency of the frequency band, and the size of this variation range is the bandwidth; D is the dispersion constant, and DM is the dispersion amount.

[0088] Step S32: Obtain the signal v propagated by the pulsar in the time domain according to the inverse of the transfer function and the Fourier transform result V(f) of the pulsar signal voltage obtained by radio telescope observation in the time domain. int The Fourier transform result V of (t) int (f).

[0089] According to formulas (2) and (12), we can know that the signal v propagated by the pulsar in the time domain is int The Fourier transform result V of (t) int (f) is:

[0090] V int (f)=V(f)H(f) -1 (13)

[0091] Among them, V(f), V int (f) and H(f) are v(t), v int (t) and the Fourier transform result of h(t), H(f)-1 is the inverse of the transfer function, and f is the observation frequency.

[0092] Step S4: Set the target bandwidth bw according to the requirements new 、Target center frequency cfreq new and target dispersion dm new , according to the target bandwidth bw new 、Target center frequency cfreq new and target dispersion dm new By adding dispersion through the inverse operation of coherent dedispersion, the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the new time domain is obtained, and IFFT is performed to obtain ultra-wideband pulsar baseband data.

[0093] The bandwidth of the original pulsar baseband data is above 100 MHz, and the target bandwidth is above 1 GHz.

[0094] Among them, according to the target bandwidth bw new 、Target center frequency cfreq new and target dispersion dm new By adding dispersion through the inverse operation of coherent dedispersion, the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the new time domain is obtained, which specifically includes:

[0095] Step S41: obtaining a target transfer function according to a target bandwidth, a target center frequency and a target dispersion amount;

[0096] Among them, the dispersion addition operation and the coherent de-dispersion operation are inverse operations. According to formula (12), the target transfer function H′(f freq +Δf) is:

[0097]

[0098] Step S42: Based on the target transfer function and the Fourier transform result V of the signal propagated by the pulsar in the time domain int (f), and obtain the Fourier transform result V′(f) of the pulsar signal voltage obtained by radio telescope observation in the new time domain.

[0099] Among them, after modifying the bandwidth, center frequency and dispersion, the Fourier transform result V′(f) of the pulsar signal voltage obtained by radio telescope observation in the new time domain is:

[0100] V′(f)=V int (f)H′(f)(15)

[0101] Where H′(f) is the target transfer function, V int(f) is the Fourier transform result of the signal propagated by the pulsar in the time domain, and V′(f) is the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the new time domain.

[0102] In addition, the method may further include step S5: packaging the ultra-wideband pulsar baseband data into a psrdada format.

[0103] Among them, when repackaging into the psrdada format, because the generated pulsar information has changed, some information in the header part of the psrdada format needs to be modified, as shown in Table 1.

[0104] Table 1 psrdada header modification information

[0105]

[0106] That is to say, when packaging into the psrdada format, the center frequency of the data header part of the psrdada format is changed to the target center frequency, the bandwidth is changed to the target bandwidth, the dispersion amount is changed to the target dispersion amount, and the sampling interval is changed to 1 / (2×target bandwidth) milliseconds.

[0107] In addition, the pulsar folding period required during pulsar processing has changed. The new pulsar folding period P new As shown in formula (16).

[0108]

[0109] Among them, P old is the folding period of the original pulsar baseband data, which is obtained from the ephemeris https: / / www.atnf.csiro.au / people / pulsar / psrcat / .

[0110] That is, step S5 also includes: calculating the folding period of the ultra-wideband pulsar baseband data finally generated according to the folding period of the original pulsar baseband data, the bandwidth of the folding period of the original pulsar baseband data and the target bandwidth. Thus, the present invention realizes the folding period correction.

[0111] The algorithm data processing speed has a lot of room for improvement and is easy to implement on GPU. The de-dispersion and addition parts are implemented on GPU, which is easy to parallelize and thus improve the processing speed.

[0112] Experimental results:

[0113] Using a bandwidth of 400MHz, a center frequency of 1382MHz, and a dispersion of 2.64476cm -3 The pulsar baseband data of PC is used as a template, such as Figure 2A-2B and Figure 3A-3B The following are the phase spectrum and pulse profile before and after de-dispersion. The ultra-wideband pulsar baseband data generation algorithm of the present invention generates a bandwidth of 3328 MHz, a center frequency of 2368 MHz, and a dispersion of 2.64476 cm -3 PC-based ultra-wideband pulsar baseband data, such as Figure 4A-4B and Figure 5A-Figure 5B As shown in the figure, they are the phase spectrum diagram and pulse profile before and after de-dispersion, which can clearly verify the correctness of the algorithm.

[0114] The above is only a preferred embodiment of the present invention and is not intended to limit the scope of the present invention. The above embodiments of the present invention can also be modified in various ways. All simple and equivalent changes and modifications made according to the claims and the description of the present invention fall within the scope of protection of the claims of the present invention. The contents not described in detail in the present invention are all conventional technical contents.

Claims

1. A method for generating ultra-wideband pulsar baseband data, characterized in that: include: Step S1: Read the original pulsar baseband data in blocks according to time to obtain data blocks; Step S2: Perform FFT on each data block to obtain frequency domain data as the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the time domain; Step S3: using the bandwidth, center frequency and dispersion of the original pulsar baseband data, a coherent de-dispersion operation is performed on the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the time domain to obtain the Fourier transform result of the signal propagated by the pulsar in the time domain; Step S4: Set the target bandwidth, target center frequency and target dispersion amount according to the requirements, add dispersion through the inverse operation of the coherent dedispersion operation according to the target bandwidth, target center frequency and target dispersion amount, obtain the Fourier change result of the pulsar signal voltage obtained by the radio telescope observation in the new time domain, and perform IFFT to obtain ultra-wideband pulsar baseband data.

2. The method for generating ultra-wideband pulsar baseband data according to claim 1, characterized in that: In step S3, a coherent de-dispersion operation is performed on the Fourier change result of the pulsar signal voltage obtained by radio telescope observation in the time domain using the bandwidth, center frequency and dispersion of the original pulsar baseband data, specifically including: Step S31: Obtain the inverse of the transfer function using the bandwidth, center frequency and dispersion of the original pulsar baseband data; Step S32: Obtaining a Fourier transform result of a signal propagated by a pulsar in the time domain according to the inverse of the transfer function and a Fourier transform result of a pulsar signal voltage obtained by radio telescope observation in the time domain; In step S4, dispersion is added by an inverse operation of the coherent de-dispersion operation according to the target bandwidth, the target center frequency and the target dispersion amount, and the Fourier change result of the pulsar signal voltage obtained by the radio telescope observation in the new time domain is obtained, which specifically includes: Step S41: obtaining a target transfer function according to a target bandwidth, a target center frequency and a target dispersion amount; Step S42: Based on the target transfer function and the Fourier transform result V of the signal propagated by the pulsar in the time domain int (f) The Fourier transform result V of the pulsar signal voltage obtained by radio telescope observation in the new time domain is obtained ′ (f).

3. The method for generating ultra-wideband pulsar baseband data according to claim 2, characterized in that: The inverse of the transfer function H(f freq +Δf) -1 for: Among them, f ref is the center frequency, Δf is the difference between the observed frequency f and the center frequency f ref The distance; f ref +Δf is the observation frequency f, and its variation range is from the minimum frequency to the maximum frequency of the band, and the size of this variation range is the bandwidth; D is the dispersion constant, and DM is the dispersion amount; The signal v propagated by a pulsar in the time domain int The Fourier transform result V of (t) int (f) is: V int (f)=V(f)H(f) -1 , Among them, V(f), V int (f) and H(f) are v(t), v int (t) and the Fourier transform result of h(t), H(f) -1 is the inverse of the transfer function, and f is the observation frequency.

4. The method for generating ultra-wideband pulsar baseband data according to claim 3, characterized in that: The target transfer function H′(f freq +Δf) is: The Fourier transform result V′(f) of the pulsar signal voltage obtained by radio telescope observation in the new time domain is: V ′ (f)=V int (f)H′(f), Where H′(f) is the target transfer function, V int (f) is the Fourier transform result of the signal propagated by the pulsar in the time domain, V ′ (f) is the Fourier transform result of the pulsar signal voltage obtained by radio telescope observation in the new time domain.

5. The method for generating ultra-wideband pulsar baseband data according to claim 1, characterized in that: In the step S1, the original pulsar baseband data is original pulsar baseband data in psrdada format and is time domain data; In step S3, the values ​​of the bandwidth, center frequency and dispersion amount of the original pulsar baseband data are specifically obtained by reading the data header part of the original pulsar baseband data in the psrdada format.

6. The method for generating ultra-wideband pulsar baseband data according to claim 1, characterized in that: The method also includes step S5: packaging the ultra-wideband pulsar baseband data into a psrdada format.

7. The method for generating ultra-wideband pulsar baseband data according to claim 6, characterized in that: When packaging into the psrdada format, the center frequency of the data header part of the psrdada format is changed to the target center frequency, the bandwidth is changed to the target bandwidth, the dispersion amount is changed to the target dispersion amount, and the sampling interval is changed to 1 / (2×target bandwidth) milliseconds.

8. The method for generating ultra-wideband pulsar baseband data according to claim 6, characterized in that: The step S5 further includes: calculating the folding period of the ultra-wideband pulsar baseband data finally generated according to the folding period of the original pulsar baseband data, the bandwidth of the folding period of the original pulsar baseband data and the target bandwidth.

9. The method for generating ultra-wideband pulsar baseband data according to claim 1, characterized in that: The bandwidth of the original pulsar baseband data is above 100 MHz, and the target bandwidth is above 1 GHz.

10. The method for generating ultra-wideband pulsar baseband data according to claim 1, characterized in that: The coherent de-dispersion operation and the inverse operation of the coherent de-dispersion operation are implemented using a GPU.

Images