CN110297233B - LFMCW array radar signal parallel pipelining processing method - Google Patents
LFMCW array radar signal parallel pipelining processing method Download PDFInfo
- Publication number
- CN110297233B CN110297233B CN201910663576.9A CN201910663576A CN110297233B CN 110297233 B CN110297233 B CN 110297233B CN 201910663576 A CN201910663576 A CN 201910663576A CN 110297233 B CN110297233 B CN 110297233B
- Authority
- CN
- China
- Prior art keywords
- processing
- lfmcw
- target
- frequency
- distance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
- G01S13/08—Systems for measuring distance only
- G01S13/32—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
- G01S13/34—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a LFMCW array radar signal parallel pipelining processing method, and aims to provide a processing method which can improve the operation efficiency, save hardware resources and reduce the system cost. The invention is realized by the following technical scheme: the digital receiving unit adopts a time-sharing sampling chip A/D integrated with an embedded programmable low noise amplifier LNA and a programmable anti-aliasing filter to receive radar array signals, adopts a digital interpolation synchronization algorithm to carry out digital processing, and then adopts a field programmable gate array FPGA to carry out fine topology on a time-space frequency three-dimensional complex signal processing algorithm so as to reduce the operation amount; and (3) constructing a time-space-frequency three-dimensional combined processing target detection mathematical model, finally decomposing and fusing algorithms of all processing modules, constructing a fine topological completely parallel flow architecture, mapping a signal processing algorithm to the FPGA, and performing parallelization, pipelining and channelization processing on the LFMCW.
Description
Technical Field
The invention relates to a parallel streamlined processing method mainly aiming at short-distance LFMCW radar, anti-collision LFMCW radar, long-distance LFMCW sky wave and ground wave radar array signals.
Background
With the development of LFMCW radar technology, chirp signals have been widely applied to the field of high-resolution radars. The chirp continuous wave radar detects information such as radial distance, speed and the like of a target by transmitting electromagnetic waves with frequency changing along with time and then detecting information such as the radial distance, the speed and the like of the target by phase difference and frequency difference between target echoes and transmitted signals, is a radar system for obtaining distance and speed information by performing frequency modulation on continuous waves, and has the advantages of small distance blind area, high distance resolution, low transmitting power, simple structure and the like, so the chirp continuous wave radar has attracted wide attention in recent years. The working distance is usually relatively small. When the action distance is increased, the number of sampling points is increased linearly under the condition of certain resolution, and in the digital processing structure, the DFT processing operand is increased according to the rule of being faster than linear, and the required memory space is also increased greatly. The LFMCW radar adopts a linear frequency modulation signal with an ultra-large time-width bandwidth product, and according to the radar signal fuzzy function theory, distance and speed coupling phenomena exist in relative motion between a target and the radar, so that the distance spectrum is fuzzy. In order to eliminate the distance measurement error caused by the distance velocity coupling, a symmetric triangle LFMCW signal can be adopted, the frequency spectrum of the same moving object beat signal generates deviation with the actual position as the symmetric axis in the up/down frequency scanning section, and the obtained frequency spectrum has the same shape. According to the two characteristics, different targets can be distinguished, pairing processing is carried out, errors caused by distance and speed coupling are eliminated, and the real distance value and speed value of the targets are measured. This method is called a beat signal frequency domain pairing method. The frequency domain pairing method is applied to point targets and distribution targets, and still has many problems:
(1) The problem of spectral overlap. When multiple targets are available, the frequency domain pairing method is easily interfered by an overlapping phenomenon generated by a target beat signal on a frequency spectrum, so that the method cannot be used;
(2) Similar spectral shape. When a plurality of targets exist, the frequency domain pairing method is easily interfered by the aspects of similar spectrum amplitude and shape of beat signals generated by different targets, and the like, so that the method is difficult to achieve the requirements. Range/doppler processing is used to extract radar target information from noise and interference. The fundamental task of LFMCW radar signal processing is to achieve detection of target echo signals and estimation of target parameters. LFMCW long-range radars usually extract target features by using several special signal processing methods: 1) Processing the arrival time of the target echo to obtain a distance parameter of the target; 2) Scanning a synthesized beam of a receiving antenna array to obtain azimuth information of a target; 3) And performing Doppler processing on the target echo to obtain the velocity information of the target. The distributed clutter caused by the moving object echo is complex in property, the echo received by the radar in the resolution unit is the synthesis of a large number of independent unit reflections, and relative motion exists in the radar, so that the synthesized echo has strong randomness. Because of the movement inside the clutter, the doppler frequency values reflected by each reflection unit are different, which causes the echo spectrum to be broadened.
The traditional LFMCW array radar obtains distance information, doppler (velocity) information and angle information by modulating a transmitting waveform and carrying out space-frequency three-dimensional coherent processing when a receiving wave is butted, and mainly comprises the contents of four aspects of distance processing, digital multi-beam forming, doppler processing and target detection. Generally, in order to obtain high range resolution, the signal bandwidth is large, and the number of range cells is large; in order to obtain higher angle resolution, the number of signal channels is more, and the number of beams required to be formed is more; in order to obtain the speed resolution capability of a low-speed target, the number of Doppler channels is large, and the processing and calculation amount of distance, direction and speed three-dimensional signals is huge, so that the traditional method of independent calculation according to four functional modules consumes more hardware resources and a large amount of calculation time, and does not meet the application scene with high real-time requirement.
Disclosure of Invention
Aiming at the problems that the calculation amount of LFMCW array radar signal processing is huge in distance dimension, angle dimension and Doppler dimension processing, more hardware resources are consumed, a large amount of calculation time is consumed, and the application scene with high real-time requirement is not met, the invention provides the LFMCW array radar signal parallel streamlined processing method which can improve the operation efficiency, save the hardware resources and reduce the system cost.
The above object of the present invention can be achieved by the following technical solutions: an LFMCW array radar signal parallel pipelining processing method has the following technical characteristics: the digital receiving unit firstly adopts a time-sharing analog-digital converter A/D sampling chip which integrates an embedded programmable low noise amplifier LNA and a programmable anti-aliasing filter, the sampling chip receives radar array signals, digital interpolation synchronization algorithm is adopted for digital processing, and then a field programmable gate array FPGA is adopted for fine topology of a time-space frequency three-dimensional complex signal processing algorithm, so that the operation amount is reduced; and (3) constructing a time-space-frequency three-dimensional combined processing target detection mathematical model, finally decomposing and fusing algorithms of all processing modules, constructing a fine topological completely parallel flow architecture, mapping a signal processing algorithm to the FPGA, and performing parallelization, pipelining and channelization processing on the LFMCW.
The invention has the following beneficial effects:
the operation efficiency is high. According to the invention, a mathematical model is constructed according to the functional performance requirements of the radar in the distance dimension, the angle dimension and the Doppler dimension, then the mathematical model is mapped to a specific FPGA chip, and then the mathematical model is gradually optimized to obtain an efficient, parallelized and streamlined algorithm structure, so that the radar signal processing is completed with the minimum hardware resources and the minimum computation delay. The defects of large resource occupation and time prolongation of the traditional FPGA signal processing of the array radar are overcome. The distance dimension, the angle dimension and the Doppler dimension are jointly designed, a time-sharing sampling chip AD8283 integrating LNA and anti-aliasing filtering is adopted to receive radar array signals, digitization processing, array radar multi-channel signal receiving and time-sharing sampling synchronization processing are conducted, then FPGA is adopted to conduct fine topology on a time-space frequency three-dimensional complex signal processing algorithm, the processing performance of parallelization, pipelining and pipelining is achieved, the operation amount is greatly reduced, and the operation efficiency is improved.
Hardware resources are saved, and system cost is reduced. The invention adopts a highly integrated A/D chip (AD 8283), and a programmable LNA and a programmable anti-aliasing filter are embedded, thereby greatly simplifying the design of the radio frequency front end of the radar; hardware multiplexing is realized through time-sharing sampling, and a digital interpolation synchronization algorithm is adopted, so that the effect of synchronous sampling is achieved, the number of A/D devices is reduced, and the system cost is reduced. The resources are saved by about 40%, and the time complexity is reduced by more than 80%.
The operation efficiency is high. The invention adopts the steps of firstly constructing a time-space-frequency processing and target detection mathematical model, then decomposing and fusing the algorithms of all processing modules, and constructing a fine-topology completely parallel flow architecture, so that the algorithms can be mapped to the FPGA for parallel and flow processing, thereby greatly improving the operation efficiency and saving the hardware resources. The method comprises the steps that one-dimensional range profiles and corresponding phase information are obtained by backing up all array element channels and all PRT synchronization data in one CPI and performing Hilbert conversion along a fast time dimension; a Taylor window and Fast Fourier Transform (FFT) are added along the spatial dimension to obtain digital multi-beam and branch data, and the target direction is distinguished; and carrying out Hamming windowing and Fast Fourier Transform (FFT) along the slow time dimension to suppress clutter of the ground objects and obtain high-resolution Doppler information. The method overcomes the defects that the traditional independent calculation method based on four functional modules consumes more hardware resources, consumes a large amount of calculation time and does not meet the application scene with higher real-time requirement.
The method adopts the Taylor window FFT along the space dimension to obtain digital multi-beam and branch data and distinguish the target direction; performing Hamming windowing and Fast Fourier Transform (FFT) along a slow time dimension to suppress clutter of the ground objects and obtain high-resolution Doppler information; and after all the sum beams detect the target information, re-extracting the backup data according to the target distance information and the Doppler information, performing differential beam forming of a specific distance unit, performing corresponding Doppler filtering, performing single pulse system and differential beam angle measurement, and acquiring accurate angle information of the target.
The invention adopts a beat signal frequency domain pairing method to eliminate the ranging error, can eliminate the influence of the distance of an insensitive target of Doppler frequency shift, the velocity coupling phenomenon and the fence effect of fast Fourier transform LFMCW on the measurement precision of the system, is insensitive to noise, and can work under a lower signal-to-noise ratio, thereby improving the speed measurement and ranging precision. Simulation results show that the method can obviously improve the LFMCW moving target detection precision under the condition of little increase of the calculation amount.
The invention is applicable to both short-range and long-range LFMCW radar systems.
Drawings
FIG. 1 is a schematic block diagram of LFMCW array radar signal parallel pipelining processing and corresponding data structure of the present invention;
FIG. 2 is a block diagram of LFMCW parallel pipeline processing according to the present invention;
fig. 3 is a diagram illustrating the structure of beam forming according to the present invention.
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments of examples. This should not be understood as limiting the scope of the above-described subject matter of the present invention to the following examples. Various substitutions and alterations according to the general knowledge and conventional practice in the art are intended to be included within the scope of the present invention without departing from the technical spirit of the present invention as described above.
Detailed Description
See fig. 1. According to the invention, a digital receiving unit firstly adopts a time-sharing analog-digital converter A/D sampling chip integrated with an embedded programmable low noise amplifier LNA and a programmable anti-aliasing filter to receive radar array signals, adopts a digital interpolation synchronization algorithm to carry out digital processing, and then adopts a field programmable gate array FPGA to carry out fine topology on a time-space-frequency three-dimensional complex signal processing algorithm so as to reduce the operation amount; and (3) constructing a time-space-frequency three-dimensional joint processing target detection mathematical model, finally decomposing and fusing algorithms of all processing modules, constructing a fine-topology complete parallel flow architecture, mapping a signal processing algorithm to the FPGA, and performing parallelization, pipelining and pipelining processing of the LFMCW.
In the parallelization, pipelining and pipelining processing, firstly, the synchronization data of all array element channels and each transmission echo in a Coherent Pulse Interval (CPI) is backed up, and then Hilbert transformation is carried out along a fast time dimension to obtain a one-dimensional range profile and corresponding phase information thereof; a Taylor window and a Fast Fourier Transform (FFT) are added along the space dimension to obtain digital multi-beam and branch data, and the target direction is distinguished; performing Hamming windowing and Fast Fourier Transform (FFT) along a slow time dimension to suppress clutter of the ground objects and obtain high-resolution Doppler information; and after the sum beam detects the target information, re-extracting Coherent Pulse Interval (CPI) backup data according to the target distance information and Doppler information, performing specific distance unit difference beam forming, performing corresponding Doppler filtering, performing single pulse system and difference beam angle measurement, and acquiring accurate angle information of the target.
Firstly, carrying out analog-to-digital (A/D) time-sharing sampling and synchronization Processing on a Beat Signal (Beat Signal) obtained by array Signal frequency mixing by a digital receiving unit to obtain L multiplied by Q multiplied by P dimensional data, wherein L is the number of antenna array element channels, Q is the number of fast time dimension sampling units, and P is the number of slow time dimension coherent Processing PRF loops, then carrying out distance data Processing (Range Processing) along the fast time dimension, and Processing to obtain L multiplied by M multiplied by P dimensional data according to the number of distance units M; taking K as the number of beams, and performing Digital Beam Forming (DBF) processing along the spatial dimension to obtain K × M × P dimensional data; using N as the number of Doppler channels, and then performing Doppler Processing (Doppler Processing) along the slow time dimension to obtain K × M × N dimensional data; and finally, performing constant false alarm and Target Processing (Detect Processing) to obtain Target information (Target Info).
And eliminating the ranging error by adopting a beat signal frequency domain pairing method. The method comprises the following specific steps: (1) determining a target. In order to eliminate the influence of the frequency spectrum side lobe of the beat signal when the target is determined, windowing is firstly carried out on the beat frequency signal, and a Hamming window or a Hanning window can be used. And determining a threshold according to the constant false alarm detection probability, and judging as a target if the constant false alarm detection probability is a section of continuous frequency spectrum continuously exceeding the threshold. And (2) selecting a pair. The beat signal frequency spectrums obtained by the upper frequency sweep section and the lower frequency sweep section of the same moving target have the same shape, and the judgment can be carried out by adopting a characteristic parameter comparison method until the beat signal LFMCW frequency spectrums of the same target are completely paired in the upper frequency sweep section and the lower frequency sweep section. And (3) determining the distance and the speed of the target.
In an alternative embodiment, a multi-channel LFMCW array radar received signal processing system is designed, and the invention is specifically explained. For the convenience of describing the present invention, the following technical terms are first defined:
definition 1: distance dimension Processing, namely fast time dimension Processing of radar echo signals; digital beam forming DBF; doppler Processing; detecting and Processing Detect Processing; beat Signal Beat Signal between continuous wave radar echo Signal and emission Signal; a radar repetition frequency PRF; a radar pulse repetition period PRT; a random access memory RAM.
Example (b):
Array element receiving signal
Wherein q is the number of samples in the fast time dimension of each period, P is the number of transmitted pulses, l is the number of array elements, A is the amplitude of echo, e is the index, j is the unit of imaginary number, fr is the difference frequency in the distance dimension, d is the distance between oscillator elements, theta is the target angle, lambda is the wavelength, fd is the Doppler frequency, T is the Doppler frequency p The method is characterized in that the method is a linear frequency modulation period, Q is a fast time dimension sampling number of each period, and L is an array element channel number.
The first exponential term in equation (1) characterizes the range difference frequency fr, which can be calculated by the following equation:
wherein, X is the frequency function of the mth distance unit when the mth array element transmits, M is the number of the distance unit, P is the number of the transmitting pulses, l is the number of the array element, Q is the distance sampling number of each period, S is zero intermediate frequency sampling data, w (Q) is a windowing coefficient used for suppressing the distance side lobe, and M is the total number of the distance units.
The second exponential term in the formula (1)
The phase difference between the ith array element and the reference array element when the characteristic receiving angle is theta can be calculated by the following formula:
digital multi-beam DBF data is obtained. Where Y is the beamforming result of the mth range bin when the kth beam is p transmit echoes, K is the beam number, w (l) is the windowing coefficient used to suppress the side lobe of the pattern, and K is the total number of beams.
In the formula (1), the third exponential term represents Doppler, doppler filtering along a slow time dimension can be calculated through the following formula, and a Doppler channel is obtainedData of
Where Z is a function of the nth Doppler channel of the mth range cell of the kth beam, n is the Doppler channel number,
for the phase factor of the p-th echo, w (p) is the windowing coefficient that suppresses the doppler side lobe.
And 2, receiving the array radar multi-channel signals and performing time-sharing sampling synchronization processing. Each A/D sampling chip of the A/D converter receives 4 paths of intermediate frequency signals (8 paths of intermediate frequency signals in total, 2 AD8283 are adopted) representing 4 array element echo signals in parallel, and the intermediate frequency signals are amplified by a Low Noise Amplifier (LNA) and filtered by an anti-aliasing filter; under the synchronization of PRF pulse, the A/D of the sampling chip of the analog-to-digital converter samples 8 paths of signals in a time-sharing way, and the sampling result is sent to the FPGA; then separating the data of each channel under the guidance of the synchronous signals, and carrying out zero filling operation, wherein the data rate of each path is improved by 4 times after zero filling; inputting the zero-filled 8-channel data into 16 paths (I and Q) of low-pass filters in parallel for filtering, and filtering out high-frequency components; and then, synchronously extracting the 8-channel filtering results output in parallel by a synchronous extractor, restoring the data rate, and storing the results in 4 RAM memories embedded in the FPGA in parallel. Therefore, the time difference among channels caused by time-sharing sampling is compensated, and the purpose of synchronization is achieved.
And 3, the distance windowing module reads the synchronous data stored in the 4 RAMs in parallel, distance windowing is carried out according to the reading beat, 8 paths of windowing results are fed into 8 fast discrete Fourier transform (FFT) modules, and distance dimensional processing results are output in parallel. Parallel data of each array element channel is fed into each beam forming module to form a complete parallel flow architecture of fine topology, so that distance dimension FFT Processing results are directly fed into a DBF Processing module (data routing and data storage are saved), distance dimension Processing and direction dimension Processing are combined into a whole, and the number of clocks required for calculation in each frequency modulation period is only 2+ (2.M) · log (2.M) + log (K) -M.
See fig. 2. In LFMCW parallel pipelining, distance dimension Processing results Range Processing #0-Range Processing # 7 are fed into 8 Complex multiply-accumulate modules (Complex multiplexer and Binary tree adders) simultaneously, 8 beam weights (weight coefficients: DBF Coefficient for angle bin # 0, #1, # 8230; # 7) are applied along direction dimension (array dimension), and results of 8 beam forming are stored in parallel in two dimensions to obtain slow time dimension and distance dimension data of 8 beams.
See fig. 3. The first beamforming complex multiply-add module is implemented as follows: the mth distance data of the nth PRT of 8 array elements is subjected to dot product operation under the parallel action of 8 complex multipliers, is subjected to addition decomposition through a delayer (register), and is accumulated on the complex multiplication result by adopting an addition tree (Binary tree adapter). The process is a pipelined operation. The remaining beams do so. Specifically, in step 3, the number of receiving array element channels is 8, and the number of doppler channels is 128. The DBF equation is re-described as:
wherein X i (m, n) is the m distance processing result of the n PRT of the ith array element channel, Y k (m, n) is the nth PRT mth distance result for the kth beam, θ k (i) Weighting the k-th beam
For example: beam 0 can be calculated as Y by 0 (m,n)＝X 0 (m,n)·θ 0 (0)+…+X 7 (m,n)·θ 7 (0)，
Weighting factor theta k (i) It can be pre-calculated and stored as a constant in the look-up table LUT, weighted with a complex multiplier. The method comprises the following steps:
and A1, directly feeding the distance dimension FFT processing result into 8 DBF modules (storage and routing are not needed), and carrying out full-pipeline pipelined processing. Entering the step A2;
a2,8 DBF modules are processed in parallel, and each module is composed of 8 complex multipliers and 1 addition tree. As shown in fig. 3.
And 4, reading a beam forming result by the Doppler processing module, performing 128-point FFT processing along a slow time dimension, and performing constant false alarm processing and target information extraction. In the step, after all the sum beams detect the target information, the difference beam forming of the specific distance unit is carried out according to the target distance information and the Doppler information, corresponding Doppler filtering is carried out, the single pulse system and the difference beam angle measurement are carried out, and the accurate angle information of the target is obtained (the difference beam only carries out the processing of limited distance and limited Doppler channels, so that the calculated amount is greatly reduced).
And eliminating the ranging error by adopting a beat signal frequency domain pairing method. The method comprises the following steps:
and B1, determining the target. Firstly, windowing is carried out on the beat frequency signal, the frequency spectrum side lobe influence of the beat signal is eliminated, then a threshold is determined according to the constant false alarm detection probability, and if a section of continuous frequency spectrum continuously exceeding the threshold exists, a target is determined. Entering a step B2;
and B2, selecting a pair. And judging that the beat signal frequency spectrums obtained by the upper and lower frequency sweep sections of the same moving target have the same shape by adopting a characteristic parameter comparison method until the beat signal LFMCW frequency spectrums of the same target are completely paired in the upper and lower frequency sweep sections. Entering a step B3;
and B3, performing pairing average on the paired distance values and speed values, and calculating the distance and speed values of the target.
Engineering verification shows that the LFMCW array radar signal parallel pipelining processing method provided by the invention can be suitable for processing all LFMCW array radar signals, and greatly improves the utilization rate of hardware resources and the operation efficiency.
Various other changes and modifications to the above-described embodiments and concepts will become apparent to those skilled in the art from the above description, and all such changes and modifications are intended to be included within the scope of the present invention as defined in the appended claims.
Claims (10)
1. An LFMCW array radar signal parallel pipelining processing method has the following technical characteristics: the digital receiving unit firstly adopts a time-sharing analog-to-digital converter A/D sampling chip integrated with an embedded programmable low noise amplifier LNA and a programmable anti-aliasing filter to receive radar array signals, adopts a digital interpolation synchronization algorithm to carry out digital processing, and then adopts a field programmable gate array FPGA to carry out fine topology on a time-space-frequency three-dimensional complex signal processing algorithm so as to reduce the operation amount; and (3) constructing a time-space-frequency three-dimensional joint processing target detection mathematical model, finally decomposing and fusing algorithms of all processing modules, constructing a fine-topology complete parallel flow architecture, mapping a signal processing algorithm to the FPGA, and performing parallelization, pipelining and pipelining processing of the LFMCW.
2. The LFMCW array radar signal parallel pipelining processing method of claim 1, wherein: in the parallelization, pipelining and pipelining processing, firstly, the synchronization data of all array element channels and each transmission echo in a Coherent Pulse Interval (CPI) is backed up, and then Hilbert transformation is carried out along a fast time dimension to obtain a one-dimensional range profile and corresponding phase information thereof; along the space dimension, a Taylor window and Fast Fourier Transform (FFT) are added to obtain digital multi-beam and branch data, and the target direction is distinguished; performing Hamming windowing and Fast Fourier Transform (FFT) along a slow time dimension to suppress clutter of the ground objects and obtain high-resolution Doppler information; and after the sum beam detects the target information, re-extracting Coherent Pulse Interval (CPI) backup data according to the target distance information and Doppler information, performing specific distance unit difference beam forming, performing corresponding Doppler filtering, performing single pulse system and difference beam angle measurement, and acquiring accurate angle information of the target.
3. The LFMCW array radar signal parallel pipelining processing method of claim 1, wherein: the digital receiving unit firstly carries out analog-to-digital (A/D) time-sharing sampling and synchronization Processing on a Beat Signal (Beat Signal) obtained by array Signal frequency mixing to obtain L multiplied by Q multiplied by P dimensional data, wherein L is the number of antenna array element channels, Q is the number of fast time dimension sampling units, and P is the number of slow time dimension coherent Processing PRF loops, then distance data Processing (Range Processing) is carried out along the fast time dimension, and the L multiplied by M multiplied by P dimensional data are obtained through Processing according to the number M of distance units; taking K as the number of beams, and carrying out digital beam forming DBF processing along the spatial dimension to obtain K multiplied by M multiplied by P dimensional data; using N as the number of Doppler channels, and then performing Doppler Processing (Doppler Processing) along the slow time dimension to obtain K × M × N dimensional data; and finally, performing constant false alarm and Target Processing (Detect Processing) to obtain Target information (Target Info).
4. The LFMCW array radar signal parallel pipelining processing method of claim 1, wherein: in order to eliminate the frequency spectrum sidelobe influence of the beat signal when the target is determined, a beat signal frequency domain pairing method is adopted to eliminate the ranging error, a distance windowing module firstly carries out windowing processing on the beat frequency signal to determine the target, a threshold is determined according to the constant false alarm detection probability, and if the target is a section of continuous frequency spectrum continuously exceeding the threshold, the target is determined; and judging that the frequency spectrums of beat signals obtained by the same moving target in the up/down frequency sweep section have the same shape by adopting a characteristic parameter comparison method, selecting pairing until the frequency spectrums of linear frequency modulation continuous waves LFMCW of the beat signals of the same target are completely paired in the up/down frequency sweep section, and determining the distance and the speed of the target.
5. The LFMCW array radar signal parallel pipelining processing method of claim 3, wherein: before subsequent processing, the M distance units perform 2M-point Fast Fourier Transform (FFT) by using the number N of Doppler channels, discard negative frequency components, and construct echo signals and fast time dimension, angle dimension and Doppler dimension processing models thereof.
6. The LFMCW array radar signal parallel pipelining processing method of claim 1, wherein: each A/D sampling chip of the A/D converter receives 4 paths of 8 intermediate frequency signals representing 4 array element echo signals in parallel, and the signals are amplified by a Low Noise Amplifier (LNA) and filtered by an anti-aliasing filter.
7. The LFMCW array radar signal parallel pipelining processing method of claim 1, wherein: under the synchronization of PRF pulse, the A/D of the sampling chip of the analog-to-digital converter samples 8 paths of signals in a time-sharing way, and the sampling result is sent to the FPGA; then separating the data of each channel under the guidance of the synchronous signal, carrying out zero filling operation, inputting the 8-channel data subjected to zero filling into a 16-channel low-pass filter in parallel for filtering, and filtering out high-frequency components; and then, synchronously extracting the 8-channel filtering results output in parallel by a synchronous extractor, restoring the data rate, and storing the results in 4 RAM memories embedded in the FPGA in parallel.
8. The LFMCW array radar signal parallel pipelining processing method of claim 1, wherein: the distance windowing module reads the synchronization data stored in the 4 RAMs in parallel, distance windowing is carried out according to the reading beat, 8 paths of windowing results are fed into the 8 fast discrete Fourier transform (FFT) modules, and distance dimensional processing results are output in parallel.
9. The LFMCW array radar signal parallel pipelining processing method of claim 7, wherein: in LFMCW parallel pipelining Processing, distance Processing results Range Processing #0-Range Processing #7 are fed into a complex multiplication accumulation module with 8 array elements, 8 wave beams are weighted along the direction array element dimension, results formed by the 8 wave beams are stored in parallel in two dimensions, and slow time dimension and distance dimension data of the 8 wave beams are obtained.
10. The LFMCW array radar signal parallel pipelining processing method of claim 9, wherein: the mth distance data of the 8 array elements, the nth PRT, is subjected to dot product operation under the parallel action of 8 complex multipliers, is subjected to additive decomposition through a delayer, and is accumulated by adopting an additive tree (Binary tree adapter).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910663576.9A CN110297233B (en) | 2019-07-23 | 2019-07-23 | LFMCW array radar signal parallel pipelining processing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910663576.9A CN110297233B (en) | 2019-07-23 | 2019-07-23 | LFMCW array radar signal parallel pipelining processing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110297233A CN110297233A (en) | 2019-10-01 |
| CN110297233B true CN110297233B (en) | 2023-02-10 |
Family
ID=68031620
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910663576.9A Active CN110297233B (en) | 2019-07-23 | 2019-07-23 | LFMCW array radar signal parallel pipelining processing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110297233B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110596671A (en) * | 2019-10-16 | 2019-12-20 | 云南大学 | Optimization processing method and system for LFMCW speed and distance measuring radar |
| CN111273271A (en) * | 2020-03-09 | 2020-06-12 | 上海无线电设备研究所 | Non-blind area distance measuring method under limited hardware resource condition |
| CN112162277B (en) * | 2020-09-29 | 2022-05-24 | 哈尔滨工业大学 | Airplane target height estimation method based on hybrid propagation high-frequency ground wave radar |
| CN113341377B (en) * | 2021-08-02 | 2021-11-05 | 网络通信与安全紫金山实验室 | Radar baseband module and radar system |
| EP4343684A1 (en) * | 2022-06-24 | 2024-03-27 | Calterah Semiconductor Technology (Shanghai) Co., Ltd | Data processing method and apparatus, and radar sensor |
| CN115656961B (en) * | 2022-12-26 | 2023-03-10 | 南京楚航科技有限公司 | OS-CFAR processing method and system based on parallel processor |
| CN115856824B (en) * | 2023-01-10 | 2023-05-09 | 电子科技大学 | Multi-phase parallel distance dimension CFAR realization method with low resource consumption |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104375140A (en) * | 2014-11-18 | 2015-02-25 | 蔚承建 | Portable through-wall radar |
| CN106054157A (en) * | 2016-07-20 | 2016-10-26 | 西安电子工程研究所 | Digital Dechirp wideband phased array radar Keystone transform algorithm |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7652617B2 (en) * | 2006-06-01 | 2010-01-26 | University Of Florida Research Foundation, Inc. | Radar microsensor for detection, tracking, and classification |
-
2019
- 2019-07-23 CN CN201910663576.9A patent/CN110297233B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104375140A (en) * | 2014-11-18 | 2015-02-25 | 蔚承建 | Portable through-wall radar |
| CN106054157A (en) * | 2016-07-20 | 2016-10-26 | 西安电子工程研究所 | Digital Dechirp wideband phased array radar Keystone transform algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110297233A (en) | 2019-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110297233B (en) | LFMCW array radar signal parallel pipelining processing method | |
| CN101915911B (en) | Space arbitrary configuration distributed SAR moving target parameter estimation method based on cancellation accumulation space-time spectrum | |
| CN106342236B (en) | Based on the airborne radar clutter suppression method of difference beam | |
| CN107976660B (en) | Missile-borne multi-channel radar ultra-low-altitude target analysis and multi-path echo modeling method | |
| CN107688178A (en) | A kind of sawtooth waveforms ranging and range rate method based on 77GHz millimetre-wave radars | |
| CN109581352B (en) | Super-resolution angle measurement system based on millimeter wave radar | |
| CN102680975B (en) | Space-based step frequency time-sharing angle measurement radar spatial non-cooperative target imaging method | |
| CN103018727A (en) | Sample-training-based non-stationary clutter suppression method of vehicle-mounted radar | |
| CN111736126B (en) | Signal processing method based on low-cost security radar | |
| CN113504522B (en) | Space-time decoupling and super-resolution angle measurement method based on random switching of transmitting antennas | |
| JP2018205174A (en) | Radar device and radar signal processing method thereof | |
| CN108535730A (en) | A kind of Doppler weather radar solution velocity ambiguity method and system | |
| CN104239683A (en) | Resolving distance-velocity ambiguity method on basis of changing signal rate | |
| CN114609623B (en) | Target detection method and device of monopulse radar and computer equipment | |
| CN111366904B (en) | CFAR detection method based on two-dimensional spectrum coherent combination | |
| CN114814817A (en) | Moving object speed deblurring method and device, electronic equipment and storage medium | |
| CN115291207A (en) | Multi-target detection method for small rotor unmanned aerial vehicle based on MIMO radar | |
| KR20200112153A (en) | Apparatus for processing signal of multi-mode radar for detecting unmanned aerial vehicle and method thereof | |
| CN109581350B (en) | Radar distance and speed measurement method and device based on time-frequency integral interpolation | |
| CN109581366B (en) | Discrete sidelobe clutter identification method based on target steering vector mismatch | |
| CN110907930B (en) | Vehicle-mounted radar target detection and estimation method and device based on angle estimation | |
| CN105319538A (en) | Accessory channel-based space-time self-adaptive clutter suppression method | |
| CN115856813A (en) | Radar target sidelobe suppression method based on APC and IARFT cascade processing | |
| Gao et al. | Static background removal in vehicular radar: Filtering in azimuth-elevation-doppler domain | |
| CN114442061A (en) | Folded clutter suppression method based on range gating and alternative inversion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |