Self-adaptive high-precision acquisition method of intelligent oscilloscope

Abstract

The invention discloses a self-adaptive high-precision acquisition method of an intelligent oscilloscope, which comprises the following steps: firstly, extracting spectrum feature distribution of a signal to be detected based on a spectrum sensing technology, then judging the type of the signal to be detected according to the spectrum feature distribution, and self-adaptively matching corresponding sampling frameworks according to different signal types; the self-adaptive dynamic configuration of the bandwidth and the sampling rate is realized; and finally, carrying out refined filtering on out-of-band noise in the acquired data by adopting adaptive filtering, and obtaining effective observation data by automatically configuring an effective observation bandwidth, so that the acquisition precision is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of oscilloscopes, and more specifically, relates to an adaptive high-precision acquisition method for an intelligent oscilloscope. Background Art

[0002] In signal testing scenarios such as radar, communication, and navigation, the signal frequency is as high as tens of GHz. Therefore, the acquisition system urgently needs a higher sampling rate and bandwidth. However, there is a conflict between bandwidth and high resolution in the acquisition system. As the bandwidth of the acquisition system increases, more noise will be introduced, resulting in a decrease in acquisition accuracy.

[0003] In traditional acquisition systems, test resources are evenly distributed over the entire bandwidth, which means that signals within the entire bandwidth range are acquired and processed. However, in most scenarios, signals often have sparse characteristics in the frequency domain. Evenly distributing bandwidth test resources will waste bandwidth resources and introduce unnecessary out-of-band noise, thereby reducing acquisition performance. On the contrary, due to the performance limitations of a single analog-to-digital converter, acquisition systems usually use time interleaving (TI) technology, frequency interleaving (FI) technology, quantization interleaving (QI) technology, asynchronous time interleaving (ATI) technology, and digital bandwidth interleaving (DBI) technology to improve the sampling rate, bandwidth, and resolution of the acquisition system. However, these interleaving technologies can also introduce non-ideal errors, thereby affecting acquisition performance. In order to make full use of hardware resources to acquire signals within the effective bandwidth and improve acquisition accuracy, people have begun to study reconfigurable acquisition architectures in recent years. Tektronix uses reconfigurable technology in the 6 Series mixed-domain acquisition system to achieve higher acquisition accuracy with a smaller sampling rate and analog bandwidth. These reconfigurable technologies realize the reconstruction of bandwidth and sampling rate in the acquisition system and improve acquisition accuracy.

[0004] However, the reconfigurable technology described above still has several problems. The first problem is that it is impossible to adaptively match signal characteristics to achieve automatic configuration of the acquisition system. Currently, the acquisition configuration needs to be manually set according to different test scenarios. The second problem is that the adjustment of the system acquisition configuration and the final test results depend on the proficiency of the tester, which limits the acquisition performance. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an adaptive high-precision acquisition method for an intelligent oscilloscope. The effective spectrum feature distribution of the signal is extracted based on spectrum sensing technology. The oscilloscope then adopts a multi-parameter adaptive matching sampling architecture to achieve adaptive dynamic configuration of bandwidth, sampling rate and resolution. At the same time, adaptive filtering is used to refine the out-of-band noise and improve the acquisition accuracy, which is convenient for effective observation of bandwidth.

[0006] To achieve the above-mentioned object of the invention, the present invention provides an adaptive high-precision acquisition method for an intelligent oscilloscope, characterized in that it comprises the following steps:

[0007] (1) The signal source generates an input signal and inputs it to the intelligent oscilloscope;

[0008] (2) In the intelligent oscilloscope, the effective spectrum interval of the input signal is extracted in real time through the sparse Fourier transform loaded in the FPGA to obtain the effective spectrum feature distribution of the input signal;

[0009] (3) judging the type of the input signal according to the effective spectrum characteristic distribution of the input signal, if the input signal is a broadband signal, proceeding to step (4); if the input signal is a sparse signal, proceeding to step (5);

[0010] (4) Collect the input signal in full bandwidth mode;

[0011] (4.1) Divide the input signal into four sub-band signals, where the bandwidths of the four sub-band signals are 0-6 GHz, 6-11 GHz, 11-16 GHz, and 16-20 GHz, respectively;

[0012] (4.2) When the four sub-band signals are subjected to analog mixing, the frequencies of the corresponding analog local oscillator signals are set to 0 GHz, 11.5 GHz, 10.5 GHz and 14.5 GHz respectively; the analog local oscillator source generates analog local oscillator signals of corresponding frequencies, and then performs analog mixing with the corresponding sub-band signals;

[0013] (4.3) Generate a digital local oscillator signal with the same frequency as the analog local oscillator signal through an adjustable numerically controlled oscillator NCO; then digitally mix the analog signal after analog mixing with the digital local oscillator signal to obtain a difference frequency signal and a sum frequency signal after digital mixing;

[0014] (4.4) The digital mixed signal is filtered through an adaptive anti-image filter to remove the difference frequency signal introduced by the up-conversion, and then the filtered signal is spliced, and the spliced ​​signal is sent to the host computer for display;

[0015] (5) Collect the input signal in intelligent acquisition mode;

[0016] (5.1) Set four sub-bands with bandwidths of 0-6 GHz, 6-11 GHz, 11-16 GHz and 16-20 GHz, and each sub-band is equipped with four low-pass filters with bandwidths of 500 MHz, 2 GHz, 4 GHz and straight-through;

[0017] (5.2) According to the effective spectrum characteristic distribution of the input signal, determine which subbands the input signal falls into, then extract the effective spectrum distribution of the corresponding subband, and extract the effective frequency band interval according to the boundary value of the effective spectrum distribution, and finally read the lowest frequency f of the effective frequency band interval low Calculate the analog local oscillator frequency f LO ;

[0018] (5.3) By controlling the clock division ratio of the phase-locked loop, the analog local oscillator source generates the corresponding frequency f LO The analog mixer performs analog mixing on the input signal and the analog local oscillator signal to obtain an intermediate frequency signal;

[0019] (5.4) Select a low-pass filter with a corresponding bandwidth according to the effective frequency band interval, and perform low-pass filtering on the intermediate frequency signal;

[0020] (5.5), the host computer simulates the local oscillator frequency f LO Generate the coefficients of the digital local oscillator and then send them to the adjustable numerically controlled oscillator NCO; the adjustable numerically controlled oscillator NCO generates a frequency based on the coefficients of the digital local oscillator and the analog local oscillator frequency f LO The same digital local oscillator signal;

[0021] (5.6) Performing real-time digital mixing of the low-pass filtered signal and the digital local oscillator signal to obtain a difference frequency signal including the digitally mixed signal and the original input signal;

[0022] (5.7) The host computer generates adaptive anti-image filter coefficients according to the spectrum distribution boundary value and effective frequency band interval falling into the sub-band, and the local oscillator frequency of the corresponding sub-band, and then sends them to the FPGA; the FPGA generates a real-time adaptive anti-image filter according to the sent coefficients; finally, the real-time adaptive anti-image filter is used to filter out the difference frequency signal and other noise after the real-time digital mixing, and retain the original input signal;

[0023] (5.8) After the signals falling into each sub-band are processed by (5.2) to (5.7), they are sent to the processing board for data splicing, and then the spliced ​​signals are adaptively filtered and finally sent to the host computer for display.

[0024] The object of the invention of the present invention is achieved in this way:

[0025] The invention discloses an adaptive high-precision acquisition method for an intelligent oscilloscope. The method first extracts the spectrum characteristic distribution of a signal to be measured based on spectrum sensing technology, then determines the type of the signal to be measured according to the spectrum characteristic distribution, and adaptively matches the corresponding sampling architecture according to different signal types to realize adaptive dynamic configuration of bandwidth and sampling rate. Finally, adaptive filtering is used to refine and filter out-of-band noise in the collected data, and effective observation data is obtained by automatically configuring the effective observation bandwidth, thereby improving the acquisition accuracy.

[0026] At the same time, the adaptive high-precision acquisition method of an intelligent oscilloscope of the present invention also has the following beneficial effects:

[0027] (1) The present invention adaptively configures the acquisition architecture according to the real-time spectrum characteristics of the signal to be tested, while the traditional acquisition system evenly distributes the test resources over the full bandwidth, which means that the signal is acquired and processed within the full bandwidth. Therefore, evenly distributing the test resources will waste bandwidth, and out-of-band noise will also reduce the acquisition performance.

[0028] (2) The present invention automatically configures the effective observation bandwidth according to the real-time spectrum characteristics of the signal to be measured, reduces the introduced noise, and improves the acquisition accuracy. Compared with the traditional full-bandwidth mode, it is more flexible, achieves a higher utilization rate of hardware resources, and can adapt to the problem of insufficient accuracy of current diverse signal types.

[0029] (3) After the present invention reduces the signal frequency by simulating the local oscillator, under the condition of the same ADC resources, more points are sampled, the signal is more accurate, the load of the ADC is reduced, and it is more convenient to perform digital signal processing. In addition, the low-frequency signal is easy to filter using a low-pass filter.

[0030] (4) The present invention uses real-time adaptive filtering to perform fine filtering on out-of-band noise to further reduce the noise, while other filtering methods cannot filter out the noise in the invalid signal noise band.

[0031] (5) The anti-imaging filter, the analog phase-locked loop generated local oscillator signal and the NCO generated local oscillator signal used in the present invention are all adaptive and generated in real time without restriction.

[0032] (6) Traditional precision improvement is based on post-acquisition processing and is not real-time. However, the present invention can process in real time upon signal input without reducing acquisition and processing performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flow chart of an adaptive high-precision acquisition method of an intelligent oscilloscope of the present invention;

[0034] Figure 2 Noisy input signal;

[0035] Figure 3 Spectrum of the input signal with noise;

[0036] Figure 4 Signal after intelligent collection; DETAILED DESCRIPTION

[0037] The specific implementation of the present invention is described below in conjunction with the accompanying drawings so that those skilled in the art can better understand the present invention. It should be noted that in the following description, when the detailed description of known functions and designs may dilute the main content of the present invention, these descriptions will be omitted here.

[0038] Example

[0039] Figure 1 The present invention discloses a flow chart of an adaptive high-precision acquisition method for an intelligent oscilloscope.

[0040] In this embodiment, if Figure 1 As shown, the present invention is an adaptive high-precision acquisition method for an intelligent oscilloscope, comprising the following steps:

[0041] (1) The signal source generates an input signal and inputs it to the intelligent oscilloscope;

[0042] (2) In the intelligent oscilloscope, the effective spectrum interval of the input signal is extracted in real time through the sparse Fourier transform loaded in the FPGA to obtain the effective spectrum feature distribution of the input signal;

[0043] (3) judging the type of the input signal according to the effective spectrum characteristic distribution of the input signal, if the input signal is a broadband signal, proceeding to step (4); if the input signal is a sparse signal, proceeding to step (5);

[0044] In this embodiment, after the FPGA obtains the real-time spectrum characteristic information of the signal, it can use the spectrum characteristic information to determine whether the oscilloscope adopts the full bandwidth mode or the intelligent acquisition mode, and configure hardware resources according to different modes.

[0045] (4) Collect the input signal in full bandwidth mode;

[0046] (4.1) Divide the input signal into four sub-band signals, where the bandwidths of the four sub-band signals are 0-6 GHz, 6-11 GHz, 11-16 GHz, and 16-20 GHz, respectively;

[0047] (4.2) When the four sub-band signals are subjected to analog mixing, the frequencies of the corresponding analog local oscillator signals are set to 0 GHz, 11.5 GHz, 10.5 GHz and 14.5 GHz respectively; the analog local oscillator source generates analog local oscillator signals of corresponding frequencies, and then performs analog mixing with the corresponding sub-band signals;

[0048] In this embodiment, each sub-band corresponds to an analog local oscillator source, and each analog local oscillator source generates an analog local oscillator signal of a corresponding frequency in the corresponding sub-band. For example, the analog local oscillator source corresponding to the third sub-band generates an analog local oscillator signal with a frequency of 10.5 GHz.

[0049] (4.3) Generate a digital local oscillator signal with the same frequency as the analog local oscillator signal through an adjustable numerically controlled oscillator NCO; then digitally mix the analog signal after analog mixing with the digital local oscillator signal to obtain a difference frequency signal and a sum frequency signal after digital mixing;

[0050] (4.4) The digital mixed signal is filtered through an adaptive anti-image filter to remove the difference frequency signal introduced by the up-conversion, and then the filtered signal is spliced, and the spliced ​​signal is sent to the host computer for display;

[0051] (5) Collect the input signal in intelligent acquisition mode;

[0052] (5.1) Set four sub-bands with bandwidths of 0-6 GHz, 6-11 GHz, 11-16 GHz and 16-20 GHz, and each sub-band is equipped with four low-pass filters with bandwidths of 20 MHz, 1 GHz, 2 GHz and straight-through;

[0053] (5.2) According to the effective spectrum characteristic distribution of the input signal, determine which subbands the input signal falls into, then extract the effective spectrum distribution of the corresponding subband, and extract the effective frequency band interval according to the boundary value of the effective spectrum distribution, and finally read the lowest frequency f of the effective frequency band interval low Calculate the analog local oscillator frequency f LO ;

[0054] Among them, the analog local oscillator frequency f LO The calculation method is:

[0055] First, the spectrum characteristics of the input signal are used to determine which sub-bands the input signal falls into. If the spectrum of the input signal falls into the third and fourth sub-bands, the spectrum distribution of the third and fourth sub-bands is extracted, and the effective frequency band interval is extracted according to the boundary value of the spectrum distribution. Finally, the lowest frequency f of the effective frequency band interval is read. low In this embodiment, the third and fourth sub-bands correspond to the lowest frequency f of the effective frequency band. low, 6GHz and 11GHz respectively, and then substitute them into the following formulas to calculate the analog local oscillator frequencies f of the third and fourth sub-bands respectively. LO ;

[0056]

[0057] (5.3) By controlling the clock division ratio of the phase-locked loop, the analog local oscillator source generates the corresponding frequency f LO The analog mixer performs analog mixing on the input signal and the analog local oscillator signal to obtain an intermediate frequency signal;

[0058] In this embodiment, for example, if the frequency spectrum of the input signal falls into the third and fourth sub-bands, the analog local oscillator frequencies f corresponding to the third and fourth sub-bands are calculated according to the frequency f of the input signal. LO ; The analog local oscillator sources corresponding to the 3rd and 4th sub-bands generate the corresponding frequencies f LO The analog local oscillator signal of the third sub-band is then mixed with the analog local oscillator frequency corresponding to the third sub-band. Similarly, the sub-band signal falling into the fourth sub-band is mixed with the analog local oscillator frequency corresponding to the fourth sub-band. Finally, the two intermediate frequency signals are processed later.

[0059] (5.4) Select a low-pass filter with a corresponding bandwidth size according to the effective frequency band range to ensure that the bandwidth of the low-pass filter can cover the effective frequency band range and boundary of the intermediate frequency signal. Perform low-pass filtering on the intermediate frequency signal to filter out high-frequency noise and the noise caused by the channel itself due to signal conditioning;

[0060] In this embodiment, the bandwidth of the intermediate frequency signal is Δw. If Δw<500MHz, a low-pass filter with a bandwidth of 500MHz is selected; if the bandwidth of the intermediate frequency signal is 500MHz<Δw≤2GHz, a low-pass filter with a bandwidth of 2GHz is selected; if the bandwidth of the intermediate frequency signal is 2GHz<Δw≤4GHz, a low-pass filter with a bandwidth of 4GHz is selected; if the bandwidth of the intermediate frequency signal is Δw>4GHz, a low-pass filter with a bandwidth of straight-through is selected;

[0061] (5.5), the host computer simulates the local oscillator frequency f LO Generate the coefficients of the digital local oscillator and then send them to the adjustable numerically controlled oscillator NCO; the adjustable numerically controlled oscillator NCO generates a frequency based on the coefficients of the digital local oscillator and the analog local oscillator frequency f LO The same digital local oscillator signal;

[0062] (5.6) Performing real-time digital mixing of the low-pass filtered signal and the digital local oscillator signal to obtain a difference frequency signal including the digitally mixed signal and the original input signal;

[0063] (5.7) The host computer generates adaptive anti-image filter coefficients according to the spectrum distribution boundary value and effective frequency band interval falling into the sub-band, and the local oscillator frequency of the corresponding sub-band, and then sends them to the FPGA; the FPGA generates a real-time adaptive anti-image filter according to the sent coefficients; finally, the real-time adaptive anti-image filter is used to filter out the difference frequency signal and other noise after the real-time digital mixing, and retain the original input signal;

[0064] (5.8) After the signals falling into each sub-band are processed by (5.2) to (5.7), they are sent to the processing board for data splicing, and then the spliced ​​signals are digitally adaptively filtered and finally sent to the host computer for display.

[0065] In this embodiment, the sub-bands where the signal spectrum falls are selected for data splicing, and the sub-bands where no signal spectrum falls are closed, and no data splicing is performed to reduce stray noise. In addition, since the analog filter cannot further refine and filter out the noise outside the effective frequency band, a digital adaptive noise suppression filter is required, so that the noise in the invalid signal noise band can be filtered out by an adaptive bandpass filter.

[0066] Examples

[0067] Input a signal with a spectrum range of 13.2Ghz-13.4Ghz and 16.2Ghz-16.4Ghz to the oscilloscope through the signal source, such as Figure 2 As shown; the signal spectrum characteristics obtained by FPGA after sparse Fourier transform are 13.2Ghz-13.4Ghz and 16.2Ghz-16.4Ghz, as shown Figure 3 As shown, all select the intelligent acquisition mode;

[0068] In the intelligent acquisition mode, according to the characteristics of the signal spectrum distribution, select the sub-band with the signal spectrum distribution and the size of the variable local oscillator value. The signal spectrum characteristics obtained by FPGA are 13.2Ghz-13.4Ghz and 16.2Ghz-16.4Ghz. Because it is in the range of 11-16GHz, it falls into the third sub-band, and because it is in the range of 16-20Ghz, it falls into the fourth sub-band. The local oscillator signal frequency of the third sub-band is calculated to be 13.2Ghz and the local oscillator signal frequency of the fourth sub-band is 16.2Ghz. Here, the effective frequency band interval length of the signal Δw=200Mhz, so a low-pass filter with a bandwidth of 500MHz is selected.

[0069] Then, the intermediate frequency signal passes through an analog low-pass filter with a passband cutoff frequency of 500MHz to filter out high-frequency noise and the noise caused by signal conditioning in the channel itself;

[0070] The filtered analog signal is then fed to the ADC. The signal is reconstructed on the FPGA. Digital up-conversion is achieved by multiplying the intermediate frequency signal with the digital local oscillator, which is generated by an adjustable numerically controlled oscillator NCO with a frequency that is consistent with the analog local oscillator frequency f. LO The same digital local oscillator signal; the coefficients of the digital local oscillator are generated by the host computer and sent to the digital hardware. After storage, the adjustable numerically controlled oscillator NCO reads the coefficients and generates. The third sub-band local oscillator signal is 13.2Ghz and the fourth sub-band local oscillator signal is 16.2Ghz.

[0071] According to the boundary values ​​13.2Ghz, 13.4Ghz, 16.2Ghz, 16.4Ghz and LO values, that is, the third sub-band local oscillator signal is 13.2Ghz and the fourth sub-band local oscillator signal is 16.2Ghz and the effective frequency band interval is 200Mhz, the host computer designs and generates the adaptive anti-image filter coefficients, and the coefficients are sent to the FPGA, stored, and then read and generated in real time. The real-time adaptive anti-image filter filters out the difference frequency signal after digital mixing, retains the original input signal and filters out other noise. The filtered signal is sent to the processing board for data splicing. Sub-band selection is performed, and the sub-band with signal spectrum falling into is selected for data splicing, and the sub-band without signal spectrum falling into is closed, and data splicing is not performed to reduce stray noise. Since the analog filter cannot further refine and filter out the noise outside the effective frequency band, a digital adaptive noise suppression filter is required. The invalid signal noise band, i.e., the noise in the 13.4Ghz-16.2Ghz band, is filtered out by an adaptive bandpass filter to obtain Figure 4 The collected signal is then sent to the host computer for display.

[0072] Although the above describes the illustrative specific embodiments of the present invention to facilitate those skilled in the art to understand the present invention, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations using the concept of the present invention are protected.

Claims

1. An adaptive high-precision acquisition method for an intelligent oscilloscope, characterized in that: The following steps are involved: (1) The signal source generates an input signal and inputs it to the intelligent oscilloscope; (2) In the intelligent oscilloscope, the effective spectrum interval of the input signal is extracted in real time through the sparse Fourier transform loaded in the FPGA to obtain the effective spectrum feature distribution of the input signal; (3) judging the type of the input signal according to the effective spectrum characteristic distribution of the input signal, if the input signal is a broadband signal, proceeding to step (4); if the input signal is a sparse signal, proceeding to step (5); (4) Collect the input signal in full bandwidth mode; (4.1) Divide the input signal into four sub-band signals, where the bandwidths of the four sub-band signals are 0-6 GHz, 6-11 GHz, 11-16 GHz, and 16-20 GHz, respectively; (4.2) When the four sub-band signals are subjected to analog mixing, the frequencies of the corresponding analog local oscillator signals are set to 0 GHz, 11.5 GHz, 10.5 GHz and 14.5 GHz respectively; the analog local oscillator source generates analog local oscillator signals of corresponding frequencies, and then performs analog mixing with the corresponding sub-band signals; (4.3) Generate a digital local oscillator signal with the same frequency as the analog local oscillator signal through an adjustable numerically controlled oscillator NCO; then digitally mix the analog signal after analog mixing with the digital local oscillator signal to obtain a difference frequency signal and a sum frequency signal after digital mixing; (4.4) The digital mixed signal is filtered through an adaptive anti-image filter to remove the difference frequency signal introduced by the up-conversion, and then the filtered signal is spliced, and the spliced ​​signal is sent to the host computer for display; (5) Collect the input signal in intelligent acquisition mode; (5.1) Set four sub-bands with bandwidths of 0-6 GHz, 6-11 GHz, 11-16 GHz and 16-20 GHz, and each sub-band is equipped with four low-pass filters with bandwidths of 500 MHz, 2 GHz, 4 GHz and straight-through; (5.2) According to the effective spectrum characteristic distribution of the input signal, determine which subbands the input signal falls into, then extract the effective spectrum distribution of the corresponding subband, and extract the effective frequency band interval according to the boundary value of the effective spectrum distribution, and finally read the lowest frequency f of the effective frequency band interval low Calculate the analog local oscillator frequency f LO ; ( 5.3) By controlling the clock division ratio of the phase-locked loop, the analog local oscillator source generates the corresponding frequency f LO The analog mixer performs analog mixing on the input signal and the analog local oscillator signal to obtain an intermediate frequency signal; (5.4) Select a low-pass filter with a corresponding bandwidth according to the effective frequency band interval, and perform low-pass filtering on the intermediate frequency signal; (5.5), the host computer simulates the local oscillator frequency f LO Generate the coefficients of the digital local oscillator and then send them to the adjustable numerically controlled oscillator NCO; The adjustable numerically controlled oscillator NCO generates a frequency that is consistent with the analog local oscillator frequency f according to the coefficients of the digital local oscillator. LO The same digital local oscillator signal; (5.6) Performing real-time digital mixing of the low-pass filtered signal and the digital local oscillator signal to obtain a difference frequency signal including the digitally mixed signal and the original input signal; (5.7) The host computer generates adaptive anti-image filter coefficients according to the spectrum distribution boundary value and effective frequency band interval falling into the sub-band, and the local oscillator frequency of the corresponding sub-band, and then sends them to the FPGA; the FPGA generates a real-time adaptive anti-image filter according to the sent coefficients; finally, the real-time adaptive anti-image filter is used to filter out the difference frequency signal and other noise after the real-time digital mixing, and retain the original input signal; (5.8) After the signals falling into each sub-band are processed by (5.2) to (5.7), they are sent to the processing board for data splicing, and then the spliced ​​signals are adaptively filtered and finally sent to the host computer for display.

2. The adaptive high-precision acquisition method of an intelligent oscilloscope according to claim 1, characterized in that: The analog local oscillator frequency f LO The calculation method is:

Images