Multi-channel nonlinear phase signal synthesis method and system

Abstract

The invention discloses a multi-channel nonlinear phase signal synthesis method and system. The method comprises the following steps: controlling a plurality of channels to collect data according to a frequency sequence; the one-out-of-two switch is controlled to be switched to a calibration signal source, and a signal is sent to multiple channels at the same time; the accurate frequency offset of each channel is obtained, and the frequency position in a baseband after one path of signal is synthesized is calculated; obtaining the accurate time difference of each channel and the amplitude difference value of each channel; calculating a mixing cos and sin phase value table according to the frequency position of each channel in the synthetic baseband, writing the phase value table into a memory, and controlling an alternative switch to be switched to an antenna to access a real signal; and performing operation processing on each channel according to the deviation value obtained by preprocessing, and merging the processed data of each channel into a signal. According to the scheme, high-precision control of time or frequency is realized without depending on complex hardware design, and rapid synthesis of multiple paths of signals can be realized at low cost through a digital algorithm mode.

Description

Technical Field

[0001] The present invention relates to the technical field of real-time software synthesis and acquisition of multi-channel signals in non-cooperative communication, and in particular to a multi-channel nonlinear phase signal synthesis method and system. Background Art

[0002] In non-cooperative communications, frequency hopping technology is widely used to ensure the security and anti-interference of communications. With the continuous development of frequency hopping technology, the frequency hopping speed is getting faster and faster, and the frequency hopping range is getting wider and wider. In addition, for large-bandwidth communication signals, especially satellite data signals, the communication bandwidth is extremely large, which leads to the need for larger real-time processing bandwidth for acquisition and receiving equipment. Due to the limitations of ADC technology development and hardware cost considerations, multi-channel ADC acquisition hardware is generally used to synthesize a large-bandwidth signal.

[0003] The existing hardware signal synthesis methods can be roughly divided into two categories: one is to use time interleaving technology to allow multiple ADCs to alternately sample the input signal in a certain time sequence, based on precise timing control to ensure the accuracy and consistency of each ADC sampling moment, so as to accurately restore the input signal. The other is frequency division multiplexing technology, which first divides the input signal into sub-signals of different frequency bands through frequency division multiplexing, and then assigns each sub-signal to a corresponding ADC for sampling, and then restores and synthesizes each sub-signal into the original large-bandwidth signal.

[0004] The first type of method is characterized by the importance of precise timing control. It is necessary to ensure that each ADC is sampled strictly according to the set time sequence. Deviations in the task sampling time may lead to sampling data errors or signal distortion. In addition, since different ADCs may have differences in parameters such as offset and gain, these parameters need to be calibrated to ensure the accuracy and reliability of the synthesized sampling results. The second type of method is characterized by the need to reasonably design a frequency division multiplexing filter group to accurately divide the signal into appropriate frequency bands for ADC sampling, and then be able to accurately restore and synthesize the sub-signals sampled by each ADC into the original signal. Summary of the invention

[0005] In view of the above technical problems, the present invention provides a multi-channel nonlinear phase signal synthesis method and system, and provides a convenient and effective multi-channel signal software synthesis method, which can quickly synthesize multi-channel signals.

[0006] The present invention is implemented by adopting the following technical scheme: a multi-channel nonlinear phase signal synthesis method, comprising the following steps:

[0007] Step S1: Control multiple channels to collect data in frequency order, and perform rough time synchronization on the data after GPS / BD timing;

[0008] Step S2: Control the two-choice switch to switch to the calibration signal source, and send a signal to multiple channels simultaneously;

[0009] Step S3: Perform FFT calculation on the multi-channel collected signals and obtain the accurate frequency deviation of each channel by finding the maximum value, and then calculate the frequency position of the synthesized signal in the baseband by combining the real collected frequencies of each channel;

[0010] Step S4: performing correlation calculation on the multi-channel collected signals to obtain the precise time difference and amplitude difference of each channel;

[0011] Step S5: Calculate the mixing cosine and sinine phase value table according to the frequency position of each channel in the synthetic baseband and write it into the memory, and control the two-choice switch to switch to the antenna to access the real signal;

[0012] Step S6: performing calculation processing on each channel according to the deviation value obtained by preprocessing, and combining the processed data of each channel into one signal.

[0013] Specifically, the calculation of the frequency position in step S3 specifically includes:

[0014] Step S31: setting the calibrated center frequencies respectively according to the n channel bandwidths and signal frequencies;

[0015] Step S32: Start acquisition to obtain the effective frequency range of n-channel baseband data;

[0016] Step S33: After performing n-fold resampling, mixing is performed, the spectrum is shifted according to a preset frequency, and finally the data is added to complete the merging.

[0017] Specifically, the step S3 further includes:

[0018] Perform SFT with the same parameters on the multi-channel collected signals and obtain the spectrum data of each channel through maximum hold. The frequency deviation of each channel is obtained by finding the maximum value of the single tone signal sent by the signal source.

[0019] Compare the time codes in the data of each channel, preliminarily synchronize the time of each channel, and then use the relevant algorithm to obtain the precise phase difference of each channel.

[0020] Specifically, the phase difference converted into time is the precise time difference of the channels. The first channel is taken as the reference channel, and the phase differences of the remaining channels relative to the reference channel are calculated. The numerical value is the maximum correlation value obtained at different offsets for different channel data.

[0021] Specifically, the related algorithm specifically includes:

[0022] Assume that the sliding window size is n, let x and y represent two different channel data sequences, and the correlation when the window slides to the kth point of the target channel data is:

[0023]

[0024] in and is the mean value of the signal sampling window, s x and y is the standard deviation, x i and i is the i-th offset point relative to the starting position of the sliding window, γ k The value range is [-1, +1];

[0025] Assume γ pq =max{γ k ,k=0,1,2,3…}|0.8≤γ k is the best correlation point of the signal, p is the position offset of the best alignment position relative to the starting data of the target channel, q is the position offset of the reference channel sliding window relative to the starting data, and the phase difference between the reference channel and the target channel is:

[0026] Time difference

[0027] Specifically, the step S6 uses opencl to accelerate parallel computing for computing each channel, and includes the following sub-steps:

[0028] Step S61: Calculate the data sampling rate after synthesizing into one signal according to the data sampling rate R of each channel;

[0029] Step S62: up-sampling each channel data, with the sampling multiple being an integer multiple, and performing interpolation filtering without decimation;

[0030] Step S63: performing frequency mixing processing on the sampled data of each channel, and completing the frequency mixing calculation by obtaining the cosine and sinine values ​​by searching the memory table;

[0031] Step S64: Calculate the amplitude mean of the mixed data of each channel;

[0032] Step S65: Calculate the delay of each channel according to the precise phase time difference and correct the phase delay;

[0033] Step S66: Use addition to merge the sampling points of each channel.

[0034] Specifically, the amplitude mean value calculation in step S64 further includes: recalculating the data amplitude value, taking the first channel as a reference, and recalculating the data amplitude values ​​of other channels.

[0035] A multi-channel nonlinear phase signal synthesis system, comprising:

[0036] Host computer module: processes the collected data of each channel in real time according to the parameters obtained by preprocessing, and sends instructions to the digital card module and the switch module;

[0037] Digital card module: receives commands from the host computer to control the underlying AD card and signal source, and reports the collected raw data to the host computer;

[0038] Switch module: switches the calibration signal source and the actual antenna receiving the real signal according to the instructions;

[0039] Data processing module: It is set in the upper computer module, with GPU performing computational processing and CPU performing logic processing.

[0040] Specifically, the sending of the host computer module instruction includes: sending instructions to the digital card through PCIE or the network to control the signal source to complete calibration preprocessing to obtain the time difference, frequency difference and amplitude difference of each channel, and then sending instructions to switch the two-choice switch to the actual antenna to receive the real signal.

[0041] The beneficial effects of the present invention are as follows: the present invention does not rely on complex hardware design to achieve high-precision control of time or frequency, and can achieve rapid synthesis of multiple signals at low cost through digital algorithms. At the same time, it can quickly synthesize multiple signals. Compared with other synthesis methods, this method has the following advantages: (1) Compared with conventional time interleaving technology and frequency division multiplexing and other hardware ADC synthesis technologies, this method can save hardware design costs; (2) Compared with conventional software synthesis algorithms, this method has higher performance and can achieve real-time synthesis of multiple signals with bandwidths of hundreds of megabits; (3) This method can be implemented on conventional CPU+GPU multi-channel acquisition devices without the need for any dedicated processing equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0043] Figure 1 A schematic diagram of multi-channel signal synthesis in an embodiment of the present invention;

[0044] Figure 2 Schematic diagram of data processing of each channel in an embodiment of the present invention;

[0045] Figure 3 Schematic diagram of frequency deviation calculation for three-channel signal synthesis in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0047] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0048] The following is combined with Figures 1 to 3 , some embodiments of the present invention are described in detail. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0049] The present invention provides a multi-channel nonlinear phase signal synthesis system. In a first aspect, in one embodiment, a multi-channel nonlinear phase signal synthesis method comprises the following steps:

[0050] Step S1: Control multiple channels to collect data in frequency order, and perform rough time synchronization on the data after GPS / BD timing;

[0051] Step S2: Control the two-choice switch to switch to the calibration signal source, and send a signal to multiple channels simultaneously;

[0052] Step S3: Perform FFT calculation on the multi-channel collected signals and obtain the accurate frequency deviation of each channel by finding the maximum value, and then calculate the frequency position of the synthesized signal in the baseband by combining the real collected frequencies of each channel;

[0053] Step S4: performing correlation calculation on the multi-channel collected signals to obtain the precise time difference and amplitude difference of each channel;

[0054] Step S5: Calculate the mixing cosine and sinine phase value table according to the frequency position of each channel in the synthetic baseband and write it into the memory, and control the two-choice switch to switch to the antenna to access the real signal;

[0055] Step S6: performing calculation processing on each channel according to the deviation value obtained by preprocessing, and combining the processed data of each channel into one signal.

[0056] In this embodiment, the calculation of the frequency position in step S3 specifically includes:

[0057] Step S31: setting the calibrated center frequencies respectively according to the n channel bandwidths and signal frequencies;

[0058] Step S32: Start acquisition to obtain the effective frequency range of n-channel baseband data;

[0059] Step S33: After performing n-fold resampling, mixing is performed, the spectrum is shifted according to a preset frequency, and finally the data is added to complete the merging.

[0060] In this embodiment, step S3 further includes:

[0061] Perform SFT with the same parameters on the multi-channel collected signals and obtain the spectrum data of each channel through maximum hold. The frequency deviation of each channel is obtained by finding the maximum value of the single tone signal sent by the signal source.

[0062] Compare the time codes in the data of each channel, preliminarily synchronize the time of each channel, and then use the relevant algorithm to obtain the precise phase difference of each channel.

[0063] In this embodiment, the phase difference converted into time is the precise time difference of the channels. The first channel is taken as the reference channel, and the phase differences of the remaining channels relative to the reference channel are calculated. The numerical values ​​are the maximum correlation values ​​obtained at different offsets for different channel data.

[0064] In this embodiment, the relevant algorithms specifically include:

[0065] Assume that the sliding window size is n, let x and y represent two different channel data sequences, and the correlation when the window slides to the kth point of the target channel data is:

[0066]

[0067] in and is the mean value of the signal sampling window, s x and y is the standard deviation, x i and i is the i-th offset point relative to the starting position of the sliding window, γ k The value range is [-1, +1];

[0068] Assume γ pq =max{γ k ,k=0,1,2,3…}|0.8≤γ k is the best correlation point of the signal, p is the position offset of the best alignment position relative to the starting data of the target channel, q is the position offset of the reference channel sliding window relative to the starting data, and the phase difference between the reference channel and the target channel is:

[0069] Time difference

[0070] In this embodiment, step S6 uses opencl to accelerate parallel computing for computing processing on each channel, and includes the following sub-steps:

[0071] Step S61: Calculate the data sampling rate after synthesizing into one signal according to the data sampling rate R of each channel;

[0072] Step S62: up-sampling each channel data, with the sampling multiple being an integer multiple, and performing interpolation filtering without decimation;

[0073] Step S63: performing frequency mixing processing on the sampled data of each channel, and completing the frequency mixing calculation by obtaining the cosine and sinine values ​​by searching the memory table;

[0074] Step S64: Calculate the amplitude mean of the mixed data of each channel;

[0075] Step S65: Calculate the delay of each channel according to the precise phase time difference and correct the phase delay;

[0076] Step S66: Use addition to merge the sampling points of each channel.

[0077] In this embodiment, the amplitude mean value calculation in step S64 further includes: recalculating the data amplitude value, taking the first channel as a reference, and recalculating the data amplitude values ​​of other channels.

[0078] The present invention also correspondingly proposes a multi-channel nonlinear phase signal synthesis system, comprising:

[0079] Host computer module: processes the collected data of each channel in real time according to the parameters obtained by preprocessing, and sends instructions to the digital card module and the switch module;

[0080] Digital card module: receives commands from the host computer to control the underlying AD card and signal source, and reports the collected raw data to the host computer;

[0081] Switch module: switches the calibration signal source and the actual antenna receiving the real signal according to the instructions;

[0082] Data processing module: It is set in the upper computer module, with GPU performing computational processing and CPU performing logic processing.

[0083] In this embodiment, the sending of the host computer module instruction includes: sending instructions to the digital card through PCIE or the network to control the signal source to complete the calibration preprocessing to obtain the time difference, frequency difference and amplitude difference of each channel, and then sending instructions to switch the two-choice switch to the actual antenna to receive the real signal.

[0084] In a preferred embodiment, multiple signal synthesis is performed as follows Figure 1 The synthesis algorithm provided by the present invention is implemented by the following steps:

[0085] Step 1: Control multiple channels to collect data in frequency order, and perform rough time synchronization through GPS / BD timing data;

[0086] Step 2: Control the two-choice switch to switch to the calibration signal source to send a signal to multiple channels simultaneously;

[0087] Step 3: Perform FFT calculation on the multi-channel acquisition signals and obtain the accurate frequency deviation of each channel by finding the maximum value. Then, combine the actual acquisition frequency of each channel to calculate the frequency position of the synthesized signal in the baseband (for example: Figure 3 , in the case of three channels, each channel has a bandwidth of 160MHz, and the total bandwidth of the three channels is 480MHz. The signal frequency is 1000MHz and the bandwidth is 400MHz. The three channels are set to have calibrated center frequencies of 840MHz+DH1, 1000MHz+DH2, and 1160MHz+DH3. After starting the acquisition, three baseband data with a bandwidth of 160MHz can be obtained. The effective frequency range is -80MHz to 80MHz. After 3 times resampling, mixing is performed to move the spectrum at -320MHz, 0MHz, and 320MHz. Finally, the three segments of data are added together to complete the merger (DH is the frequency deviation calculated by the calibration signal source).

[0088] Step 4: Perform "correlation" calculation on the multi-channel acquisition signals to obtain the precise time difference of each channel;

[0089] Step 5: Average the amplitudes of the multi-channel collected signals to calculate the amplitude difference of each channel;

[0090] Step 6: Calculate the mixing cosine and sinine phase values ​​according to the frequency position of each channel in the synthetic baseband and write them into the memory;

[0091] Step 7: Control the two-choice switch to switch the antenna to access the real signal;

[0092] Step 8: Perform calculation processing on each channel according to the deviation value obtained by preprocessing;

[0093] Step 9: Combine the processed data from each channel into one signal.

[0094] Among them, the FFT calculation of each channel is an SFT (short-time Fourier transform) with the same parameters and the spectrum data of each channel is obtained by maximum hold. Since the signal sent by the signal source is a single-tone signal, the frequency deviation of each channel can be obtained by finding the maximum value.

[0095] Among them, by comparing the time code in the data of each channel, the time of each channel can be preliminarily synchronized, and then the precise phase difference of each channel can be obtained through the relevant algorithm.

[0096] Among them, by comparing the time code in each channel data, the time of each channel can be preliminarily synchronized, and then the accurate phase difference of each channel can be obtained through the correlation algorithm. After this phase difference is converted into time, it is the channel accurate time difference. Taking the first channel as the reference channel, the phase difference of the remaining channels relative to the reference channel is calculated. The numerical value is reflected in the maximum correlation value obtained at different offsets of different channel data. The basic correlation algorithm adopts the Pearson correlation algorithm of the sliding window. The principle is as follows:

[0097] Assume that the sliding window size is n, let x and y represent two different channel data sequences, and the correlation when the window slides to the kth point of the target channel data is in is the mean value of the signal sampling window, s x 、s y is the standard deviation, x i ,y i is the i-th offset point relative to the starting position of the sliding window, γ k The value range is [-1, +1], let γ pq =max{γ k ,k=0,1,2,3…}|0.8≤γ k is the best correlation point of the signal, at this time p is the position offset of the best alignment position relative to the starting data of the target channel, q is the position offset of the reference channel sliding window relative to the starting data, and the phase difference between the reference channel and the target channel can be obtained Time difference This phase difference / time difference will be used for subsequent correction. In actual operation, the reference channel usually takes a fixed window size of data and keeps it unchanged, and only the target channel data window slides. If the best correlation position that meets the conditions is not found after sliding for a sufficient length, the reference channel window slides back a specified distance and recalculates with the target channel data. The phase difference of the remaining channels is obtained by performing the above correlation calculations on the reference channel and different channels.

[0098] Among them, step 8 uses opencl to accelerate the parallel operation of each channel. Figure 2 The following is a schematic diagram of data processing for each channel. The specific steps can be described as follows:

[0099] Step a: Calculate the data sampling rate after synthesizing one signal according to the data sampling rate R of each channel (considering the case where the sampling rates of each channel are the same, the sampling rate after synthesis is N*R, where N is the number of channels);

[0100] Step b: upsample the data of each channel. If the sampling multiple is an integer multiple, only interpolation filtering is required without decimation.

[0101] Step c: Mix the sampled data of each channel. By looking up the memory table, the cosine and sinine values ​​can be quickly obtained to complete the mixing calculation.

[0102] Step d: Calculate the amplitude mean of the mixed data of each channel;

[0103] Step e: Calculate the delay of each channel according to the precise phase time difference and correct the phase delay;

[0104] Step f: Use addition to merge the sampling points of each channel.

[0105] The data amplitude value is recalculated based on the first channel to recalculate the data amplitude values ​​of other channels (based on the mean value calculation ratio).

[0106] This embodiment mainly includes two parts: a digital card and a host computer. The digital card mainly receives instructions from the host computer to complete the control of the underlying AD card and the signal source and report the collected raw data to the host computer. Signal synthesis is mainly completed by the host computer. The host computer sends instructions to the digital card through PCIE or the network to control the signal source to complete the calibration preprocessing to obtain the time difference, frequency difference and amplitude difference of each channel, and then sends instructions to switch the two-choice switch to the actual antenna to receive the real signal. The host computer processes the collected data of each channel in real time according to the parameters obtained by the preprocessing. The calculation level processing is handled by the GPU, and the CPU performs logic processing.

[0107] The following briefly describes an implementation of the signal synthesis method using a general three-channel radio digital receiver and a general PC as an example:

[0108] The receiver channel bandwidth is 160MHz, corresponding to a sampling rate of 204.8MSps. The three channels have a collection center frequency of 1000MHz and a bandwidth of 480MHz. The frequency ranges of the three channels are (840+df1):160, (1000+df2):160, (1160+df3):160, respectively. df1, df2, and df3 are frequency deviations calculated by correlation of the three channels. The bandwidth of the three channels after synthesis is 480MHz, and the sampling rate is 204.8*3=614.4MSps. After the host computer receives the three-channel reported data through PCIE, it performs 3 times upsampling and determines the three channel mixing positions in the range of -fs / 2 to fs / 2: -320MHz, 0MHz, and 320MHz. The three channels are mixed, and the amplitude values ​​are recalculated and then the three channels are added to output the synthesized data.

[0109] The multi-channel signal combining method proposed in the present invention proposes an algorithm that does not rely on multi-channel ADC time interleaving and frequency division multiplexing, which is suitable for realizing real-time ZC of large bandwidth signals at a relatively low cost; GPS / BD is used to provide time to multi-channel acquisition data and write time codes, and the phase and amplitude between the channels are calibrated and synchronized through the self-checking signal source structure correlation algorithm; the time required for mixing operation can be greatly reduced through fast mixing calculation of the table lookup method, multi-phase filtering, and opencl acceleration. After obtaining accurate phase difference and amplitude difference, multi-channel signal synthesis is performed quickly and accurately.

[0110] For the aforementioned embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily required by the present application.

[0111] The above embodiments describe the basic principles and main features of the present invention and the advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the changes and modifications made by those skilled in the art shall be within the scope of protection of the appended claims of the present invention without departing from the spirit and scope of the present invention.

Claims

1. A multi-channel nonlinear phase signal synthesis method, characterized in that: The following steps are involved: Step S1: Control multiple channels to collect data in frequency order, and perform rough time synchronization on the data after GPS / BD timing; Step S2: Control the two-choice switch to switch to the calibration signal source, and send a signal to multiple channels simultaneously; Step S3: Perform FFT calculation on the multi-channel collected signals and obtain the accurate frequency deviation of each channel by finding the maximum value, and then calculate the frequency position of the synthesized signal in the baseband by combining the real collected frequencies of each channel; Step S4: performing correlation calculation on the multi-channel collected signals to obtain the precise time difference and amplitude difference of each channel; Step S5: Calculate the mixing cosine and sinine phase value table according to the frequency position of each channel in the synthetic baseband and write it into the memory, and control the two-choice switch to switch to the antenna to access the real signal; Step S6: performing calculation processing on each channel according to the deviation value obtained by preprocessing, and combining the processed data of each channel into one signal.

2. A multi-channel nonlinear phase signal synthesis method as claimed in claim 1, characterized in that: The calculation of the frequency position in step S3 specifically includes: Step S31: setting the calibrated center frequencies respectively according to the n channel bandwidths and signal frequencies; Step S32: Start acquisition to obtain the effective frequency range of n-channel baseband data; Step S33: After performing n-fold resampling, mixing is performed, the spectrum is shifted according to a preset frequency, and finally the data is added to complete the merging.

3. A multi-channel nonlinear phase signal synthesis method as claimed in claim 1, characterized in that: The step S3 further comprises: Perform SFT with the same parameters on the multi-channel collected signals and obtain the spectrum data of each channel through maximum hold. The frequency deviation of each channel is obtained by finding the maximum value of the single tone signal sent by the signal source. Compare the time codes in the data of each channel, preliminarily synchronize the time of each channel, and then use the relevant algorithm to obtain the precise phase difference of each channel.

4. A multi-channel nonlinear phase signal synthesis method as claimed in claim 3, characterized in that: The phase difference converted into time is the precise time difference of the channels. The first channel is taken as the reference channel, and the phase differences of the remaining channels relative to the reference channel are calculated. The numerical value is the maximum correlation value obtained at different offsets for different channel data.

5. A multi-channel nonlinear phase signal synthesis method as claimed in claim 3, characterized in that: The relevant algorithms specifically include: Assume that the sliding window size is n, let x and y represent two different channel data sequences, and the correlation when the window slides to the kth point of the target channel data is: in and is the mean value of the signal sampling window, s x and y is the standard deviation, x i and i is the i-th offset point relative to the starting position of the sliding window, γ k The value range is [-1, +1]; Assume γ pq =max{γ k ,k=0,1,2,3…}|0.8≤γ k is the best correlation point of the signal, p is the position offset of the best alignment position relative to the starting data of the target channel, q is the position offset of the reference channel sliding window relative to the starting data, and the phase difference between the reference channel and the target channel is: Time difference 6. A multi-channel nonlinear phase signal synthesis method as claimed in claim 1, characterized in that: The step S6 performs operation processing on each channel using opencl for parallel operation acceleration, and includes the following sub-steps: Step S61: Calculate the data sampling rate after synthesizing into one signal according to the data sampling rate R of each channel; Step S62: up-sampling each channel data, with the sampling multiple being an integer multiple, and performing interpolation filtering without decimation; Step S63: performing frequency mixing processing on the sampled data of each channel, and completing the frequency mixing calculation by obtaining the cosine and sinine values ​​by searching the memory table; Step S64: Calculate the amplitude mean of the mixed data of each channel; Step S65: Calculate the delay of each channel according to the precise phase time difference and correct the phase delay; Step S66: Use addition to merge the sampling points of each channel.

7. A multi-channel nonlinear phase signal synthesis method as claimed in claim 6, characterized in that: The amplitude mean value calculation in step S64 also includes: recalculating the data amplitude value, taking the first channel as a reference, and recalculating the data amplitude values ​​of other channels.

8. A multi-channel nonlinear phase signal synthesis system, used to implement a multi-channel nonlinear phase signal synthesis method according to any one of claims 1 to 7, characterized in that: include: Host computer module: processes the collected data of each channel in real time according to the parameters obtained by preprocessing, and sends instructions to the digital card module and the switch module; Digital card module: receives commands from the host computer to control the underlying AD card and signal source, and reports the collected raw data to the host computer; Switch module: switches the calibration signal source and the actual antenna receiving the real signal according to the instructions; Data processing module: It is set in the upper computer module, with GPU performing computational processing and CPU performing logic processing.

9. A multi-channel nonlinear phase signal synthesis system as claimed in claim 8, characterized in that: The sending of the host computer module instruction includes: sending instructions to the digital card through PCIE or the network to control the signal source to complete calibration preprocessing to obtain the time difference, frequency difference and amplitude difference of each channel, and then sending instructions to switch the two-choice switch to the actual antenna to receive the real signal.

Images