Dual-channel quasi-coherent signal processing method based on magnetron pulse radar

Abstract

The invention relates to a dual-channel quasi-coherent signal processing method based on a magnetron pulse radar. Comprising the following steps: arranging two receiving channels, a detection channel and an echo channel on a radar receiver; basic data preprocessing is carried out on the two paths of original signal data; performing phase correction on a target echo signal by using a transmitting signal collected by a detection channel to obtain a plurality of groups of echo data corrected according to a pulse sequence; performing three-dimensional matrix modeling according to the pulse sequence to obtain a data block containing multiple groups of echo signal data after coherent correction; doppler processing is carried out on the obtained data blocks, and the speed measurement function of the target object is achieved. According to the invention, the signal-to-noise ratio of a target signal can be greatly improved, and the radar target detection capability is enhanced; the power is high and the size is small; the method can be used in the field of marine navigation radars and the field of low-altitude detection, realizes monitoring of low-altitude unmanned aerial vehicles, and is very wide in application range.

Description

Technical Field

[0001] The present invention relates to the technical field of radar signal processing, and specifically relates to a dual-channel quasi-coherent signal processing method based on a magnetron pulse radar. Background Art

[0002] The magnetron pulse radar is a type of radar widely used in radar systems, especially in the military, meteorological, aerospace and other fields. Since the end of the last century, this type of radar has gradually withdrawn from the military field, and its performance cannot meet the ever-changing and complex radar electronic warfare. However, it is still the mainstay in the civilian field, occupying half of the country. Its high power, low cost, and relatively simple design are deeply loved by some enterprises. It uses a magnetron to generate high-frequency pulse signals, and detects the position, speed and other physical characteristics of target objects by transmitting and receiving the reflected signals. At the same power, the cost of coherent radars is very high, and the relatively large volume is restricted in deployment. Therefore, the research on the quasi-coherent processing of such low-cost magnetron radars is of great significance.

[0003] In a magnetron pulse radar, the reception and processing of signals are one of the key factors affecting the performance of the radar system. The traditional radar signal processing method is based on single-channel signal processing. By processing with the local oscillator signal, the distance information of the target can be obtained. However, due to the inherent physical structure of the magnetron, the electromagnetic wave signals generated by it are non-coherent signals. Non-coherent signals can only be non-coherently accumulated in the subsequent signal processing process, resulting in a poor signal-to-noise ratio of the target. At the same time, due to the non-coherence between signals, the phase correlation between signals is lost, resulting in the inability to measure the speed of the target object. In "CN111913154 A", "Magnetron Radar Reception Coherent Digital Processing Method" is given. Since the structure is single-channel and the transmitted samples are obtained through traditional radio frequency coupling technology, there are problems such as coupler loss and introduction of non-linear distortion. Although the quasi-coherent reception processing of the magnetron radar can be realized in this way, there are deficiencies in terms of accuracy and stability.

[0004] Under the magnetron pulse radar system, in order to obtain a higher signal-to-noise ratio and realize the speed measurement of the target object, it is still necessary to research the reception and processing of signals. Summary of the Invention

[0005] The present invention provides a dual-channel quasi-coherent signal processing method based on a magnetron pulse radar to overcome the problems of poor target signal-to-noise ratio and inability to measure the speed of the target object existing in the prior art.

[0006] To achieve the above object, the technical solution of the present invention is: A dual-channel quasi-coherent signal processing method based on a magnetron pulse radar, specifically including the following steps:

[0007] Step 1: Set up two receiving channels on the radar receiver. One receiving channel is used to receive the transmitted signal of the magnetron radar, which is called the detection channel; the other receiving channel normally collects the target echo signal through the antenna, which is called the echo channel.

[0008] Step 2: Perform basic data preprocessing on the two sets of original signal data collected by the detection channel and the echo channel in Step 1.

[0009] Step 3: Connect the complex signals of the detection channel and the echo channel to the quasi-coherent correction module, and use the transmitted signal collected by the detection channel to perform phase correction on the target echo signal to obtain multiple sets of echo signal data corrected according to the pulse order.

[0010] Step 4: Store the multiple sets of corrected echo signal data in Step 3, and perform three-dimensional matrix modeling according to the pulse order to obtain a data block containing multiple sets of coherently corrected echo signal data.

[0011] Step 5: Perform Doppler processing on the obtained data block, process the slow time dimension of the data block, extract the Doppler frequency shift amount, and realize the speed measurement of the target object.

[0012] Further, in the above Step 2, the basic data preprocessing includes the following steps:

[0013] 2.1. Perform down-conversion processing on the transmitted signal collected by the detection channel and the echo signal collected by the echo channel in Step 1 simultaneously through the RF front-end receiving chip, and output a typical single-ended intermediate frequency signal.

[0014] 2.2. For the intermediate frequency signal output after the down-conversion processing of the RF front-end receiving chip, use an AD high-speed analog-to-digital converter for intermediate frequency sampling to convert it into an intermediate frequency digital signal.

[0015] 2.3. After the signal data obtained through the AD high-speed analog-to-digital conversion is output, it will be divided into the I channel and the Q channel. There are a total of 4 digital signals for the detection channel and the echo channel. Perform band-pass filtering on them to filter out noise interference in other frequency bands.

[0016] 2.4. After passing through the band-pass filter, combine the corresponding I channel and Q channel in the detection channel into one complex signal. Similarly, combine the corresponding I channel and Q channel in the echo channel into another complex signal.

[0017] Further, the specific steps of the above Step 3 are as follows:

[0018] 3.1. Perform Fourier transform on the recombined complex signal in the detection channel, and simultaneously perform Fourier transform on the recombined complex signal in the echo channel.

[0019] 3.2. Take the conjugate of the digital complex signal in the detection channel after Fourier transform, and perform a complex multiplication operation with the digital signal in the echo channel after Fourier transform;

[0020] 3.3. Perform an inverse Fourier transform on the data after complex multiplication to obtain the target echo data after coherent correction.

[0021] Furthermore, the mathematical model expression of the echo signal data output in the above 3.3 is as follows:

[0022]

[0023] Among them, E R (t) is the digital complex signal in the detection channel; E T (t) is the digital complex signal in the echo channel. After convolution of the two, the same random initial phase contained in both can be eliminated. Therefore, no matter how the phase changes during transmission, the changing trend of the output phase is the same and very stable, which plays a role in phase correction for the signal.

[0024] Furthermore, when performing Doppler processing in the above step five, first change the arrangement method of the matrix data, then arrange them in sequence according to the slow time dimension, and finally pass the rearranged data through a Doppler filter to obtain the Doppler frequency shift through spectrum analysis.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. The method of the present invention is based on a radar receiver with two receiving channels, which can be directly manufactured or, on the basis of basically not changing the original radar hardware facilities, add a receiving antenna, that is, there are two receiving channels on the radar receiver, and these two receiving channels are the detection channel and the echo channel respectively. By using a dual-channel to receive signals for a magnetron pulse radar, each pulse signal emitted by the transmitter is collected in real time as a reference signal. By performing convolution processing on the reference signal and the target echo signal, quasi-coherent correction is performed on the received non-coherent signal. By monitoring the transmitted signal, the random phase in the echo signal is eliminated, effectively eliminating the random phase noise in the echo signal, aligning the phases, making the pulse signals coherent, greatly improving the signal-to-noise ratio of the target signal, and enhancing the radar target detection ability.

[0027] 2. The present invention uses a dual-channel quasi-coherent signal processing method for magnetron pulse radar to perform coherent correction on the signals of non-coherent radar. This technology realizes the moving target indication (MTI) and pulse Doppler (PD) processing capabilities at the coherent radar level on a magnetron radar for the first time. The provided method can help non-coherent magnetron radars achieve moving target detection, and can enhance the detection effect of magnetron radars on the surrounding environment by coherent accumulation to improve the signal-to-noise ratio, greatly enhancing the target detection ability of magnetron radars.

[0028] 3. The processing method of the present invention is compatible with the existing radar hardware architecture, and the transformation cost is very low. It can achieve the performance level of coherent radar at a relatively low cost, effectively improving the economy and performance of the radar system. The price cost of coherent radar is very high under the same power, so the magnetron with low cost, high power and small volume has extremely high power cost performance.

[0029] 4. It can not only be used in the field of marine navigation radar to improve the detection performance of the surrounding environment, but also roughly estimate the speed of nearby moving targets. It can also be used in the field of low-altitude detection to realize the monitoring of low-altitude unmanned aerial vehicles, and has a very wide range of applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a flowchart of the dual-channel quasi-coherent signal processing method of the present invention;

[0031] Figure 2 is a schematic diagram of the quasi-coherent correction structure of the present invention;

[0032] Figure 3 is a schematic diagram of the measured transmitted signal of a magnetron pulse radar;

[0033] Figure 4 is a schematic diagram of the measured echo signal of a magnetron pulse radar;

[0034] Figure 5 is a schematic diagram of the principle structure of the convolution processing of the present invention;

[0035] Figure 6 is a schematic diagram of the data processing process of the quasi-coherent correction of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0037] Embodiment

[0038] See Figure 1, A signal receiving and processing device in a conventional magnetron pulse radar, including a radio frequency front-end chip, AD intermediate frequency sampling, and a single channel composed of two band-pass filters. On the basis of the original radar hardware facilities, the present invention adds a channel identical to the original channel, that is, adds a receiving antenna, and also includes a radio frequency front-end chip, AD intermediate frequency sampling, and two band-pass filters.

[0039] A dual-channel quasi-coherent signal processing method based on a magnetron pulse radar provided by the present invention performs a convolution operation on the data in the detection channel and the echo channel, thereby realizing the quasi-coherent correction of the echo signal. The memory of the present invention is used to store the corrected echo signal data, and then the subsequent signal processing is completed by reading the memory.

[0040] Specifically, in this embodiment, the processing object is specifically a magnetron marine navigation radar. Further, the processing object is echo signal data. See Figure 2 , The present invention adopts a dual-channel quasi-coherent signal processing method based on a magnetron pulse radar for a specific object, which specifically includes the following steps:

[0041] Step 1: On the basis of the original radar receiver, add another receiving channel to receive the transmitted signal of the magnetron radar, which is called the detection channel. The detection channel is used to collect each pulse signal emitted by the transmitter. By performing a convolution process on the reference signal and the target echo signal, the random phase noise in the echo signal can be effectively eliminated, the phases can be aligned, and the signal coherence effect can be achieved. Another receiving channel normally collects the target echo signal through the antenna, which is called the echo channel.

[0042] Refer to Figures 3 to 4 , The expressions corresponding to a set of pulse data received by the detection channel and the echo channel are respectively:

[0043]

[0044] Here, U T represents the data in the detection channel, and U R represents the data in the echo channel. represents the random phase vector before each pulse.

[0045] The transmission of a set of pulse data and its corresponding echo will both contain the same vector This is an inevitable defect in the hardware structure of the magnetron radar to generate high-power microwaves. Since the vector before each set of pulse data is different and is a random variable, this vector is called the random phase of the pulse. The present invention effectively eliminates the influence brought by this vector in the subsequent steps through the re-acquisition of this vector by the detection channel.

[0046] Step 2. Perform basic data preprocessing on the two-channel original signal data collected in the detection channel and the echo channel in Step 1, including the following steps:

[0047] 2.1. Simultaneously perform down-conversion processing on the transmitted signal collected in the detection channel in Step 1 and the echo signal collected in the echo channel through the RF front-end receiving chip, and output a typical single-ended intermediate frequency signal;

[0048] The two intermediate frequency signals after simultaneous down-conversion are:

[0049]

[0050]

[0051] Here, G t 、G r are the transmission gains of the receiving detection channel and the receiving echo channel, A represents the amplitude of the transmitted signal, B represents the signal propagation attenuation coefficient, and ω d represents the Doppler frequency shift of the target.

[0052] 2.2. For the intermediate frequency signal output after the down-conversion processing by the RF front-end receiving chip, use an AD high-speed analog-to-digital converter for intermediate frequency sampling and convert it into an intermediate frequency digital signal;

[0053] For the intermediate frequency signal output after the down-conversion processing by the RF front-end receiving chip in the echo channel, use an AD high-speed analog-to-digital converter for intermediate frequency sampling and convert it into an intermediate frequency digital signal; similarly, for the intermediate frequency signal output after the down-conversion processing by the RF front-end receiving chip in the detection channel, use an AD high-speed analog-to-digital converter for intermediate frequency sampling and convert it into an intermediate frequency digital signal;

[0054] In the detection channel, by calculating the number of intermediate frequency digital signal points corresponding to the pulse width, perform a sampling operation, only take the points within the previous pulse width, and fill the subsequent remaining points with zeros, which can not only minimize the influence of noise on the data processing target detection result to the greatest extent, but also reduce the complexity of the operation.

[0055] 2.3. The signal data output after the AD high-speed analog-to-digital conversion will be divided into the I channel and the Q channel. The additional detection channel and the echo channel of the original receiver will have a total of 4 digital signals at this time. Perform band-pass filtering on them to filter out noise interference in other frequency bands;

[0056] 2.4. After passing through the band-pass filter, combine the corresponding I channel and Q channel in the detection channel into one complex signal. Similarly, combine the corresponding I channel and Q channel in the echo channel into another complex signal.

[0057] Step 3: Connect the complex signals of the detection channel and the echo channel to the quasi-coherent correction module, and use the transmitted signal collected by the detection channel to perform phase correction on the target echo signal to obtain multiple groups of echo signal data corrected according to the pulse order. The specific steps are as follows:

[0058] 3.1 Perform Fourier transform on the re-synthesized complex signal in the detection channel, and also perform Fourier transform on the re-synthesized complex signal in the echo channel at the same time sequence.

[0059] 3.2 Take the conjugate of the digital complex signal in the detection channel after Fourier transform, and perform complex multiplication operation with the digital signal in the echo channel after Fourier transform.

[0060] 3.3 Perform inverse Fourier transform on the data after complex multiplication to obtain the target echo data after coherent correction:

[0061] Here, the convolution operation is performed on two sets of data in the two channels through the signal expression after convolution processing; the mathematical model expression of the corrected echo signal data (that is, the result after coherent correction for a single pulse) is as follows:

[0062]

[0063] Among them, E R (t) is the digital complex signal in the detection channel; E T (t) is the digital complex signal in the echo channel; after convolution of the two, the same random initial phase contained in both can be eliminated. Therefore, no matter how the phase changes during transmission, the output phase change trend is the same and very stable, which plays a role in phase correction of the signal.

[0064] In this Step 3, the convolution of the two signals is realized by the method of frequency domain calculation. First, since the transmission and echo of the same signal source have strong autocorrelation, the random phase error between the signals generated by the non-coherent magnetron radar can be effectively eliminated by convolution; second, the method of converting to frequency domain calculation can reduce the computational complexity, and a ten-fold speed improvement can be obtained when processing long sequences. Moreover, the frequency domain dot product operation naturally supports parallel architectures, and a hundred-fold acceleration can be achieved with the help of modern computing units.

[0065] See Figure 5 , the present invention converts to the frequency domain through Fourier transform, and uses the equivalence of frequency domain multiplication to time domain convolution to realize the convolution processing of signal data, and this can greatly reduce the computational amount of time domain convolution and better complete the process of quasi-coherent correction. The present invention eliminates random vectors by convolution to achieve the purpose of phase correction. The echo correction data within one pulse repetition period is recorded as a set of data, and N sets of corrected data are recorded.

[0066] Step 4: Store the multi-group corrected echo signal data in Step 3, and perform three-dimensional matrix modeling according to the pulse order to obtain a data block containing multi-group coherent corrected echo signal data;

[0067] Step 5: Perform Doppler processing on the obtained data block, extract the Doppler frequency shift amount, and realize the speed measurement of the target object:

[0068] When performing Doppler processing, first change the arrangement mode of the matrix data, then arrange them in sequence according to the slow time dimension, and finally pass the rearranged data through a Doppler filter, and obtain the Doppler frequency shift through spectrum analysis.

[0069] Refer to Figure 6 , longitudinally arrange the N groups of corrected data in order, construct a two-dimensional data matrix to form a data block, where the horizontal direction represents the fast time and the vertical direction represents the slow time; the corrected signal has a certain degree of quasi-coherence, and coherent accumulation can greatly improve the signal-to-noise ratio of the target signal and enhance the perception ability of the surrounding environment.

[0070] A more detailed description of Step 5 is as follows:

[0071] Perform FFT on the longitudinal data of the data block matrix, that is, in the direction of the slow time axis, to obtain a Doppler spectrogram for extracting the frequency shift magnitude caused by the movement of the target object;

[0072] Obtain the speed information through the mathematical relationship between the Doppler frequency shift and the speed. Normally, the echo signals of a non-coherent radar do not have coherence, and its Doppler frequency domain is chaotic, and it is impossible to estimate the speed of the target. This step can be achieved through the coherent correction method of this method.

[0073] Perform a constant false alarm rate detection process on the two-dimensional Doppler frequency domain to complete the realization of target ranging and speed measurement of a magnetron non-coherent radar under the dual-channel quasi-coherent processing method.

[0074] The following analyzes the improvement effect of coherent accumulation of the target signal after coherent correction through the detection effect of a magnetron navigation radar on the surrounding environment (stationary target). At the same time, through the detection results of the speed of a moving target by a magnetron navigation radar, analyze whether the data after coherent correction can perform Doppler processing and whether the speed measurement is accurate. Multiple experimental data show that the effect of coherent accumulation after quasi-coherent processing can reach more than 96% of that of a fully coherent radar, and the detected values of moving targets are also within the speed resolution of a Doppler pulse radar.

[0075] On the premise that the original basic hardware of the magnetron radar remains unchanged, this method adds a detection channel, including a receiving antenna, a radio frequency down-conversion chip, an AD intermediate frequency sampling chip, a band-pass filter, etc., to correct the problem of incoherence of the phase of the echo signal in the echo channel through this detection channel.

[0076] The foregoing are only the preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A dual-channel quasi-coherent signal processing method based on a magnetron pulsed radar, characterized in that: It includes the following steps: Step 1: Set two receiving channels on the radar receiver. One receiving channel is used to receive the transmitted signal of the magnetron radar, which is called the detection channel; the other receiving antenna normally collects the target echo signal, which is called the echo channel. Step 2: Perform basic data preprocessing on the two sets of original signal data collected by the detection channel and the echo channel in Step 1. Step 3: Connect the complex signals of the detection channel and the echo channel to the quasi-coherent correction module, and use the transmitted signal collected by the detection channel to perform phase correction on the target echo signal to obtain multiple sets of echo data corrected according to the pulse order. Step 4: Store the multiple sets of corrected echo signal data in Step 3, and perform three-dimensional matrix modeling according to the pulse order to obtain a data block containing multiple sets of coherently corrected echo signal data. Step 5: Perform Doppler processing on the obtained data block, extract the Doppler frequency shift amount, and realize the speed measurement function of the target object.

2. The dual-channel quasi-coherent signal processing method based on a magnetron pulse radar according to claim 1, wherein: In the above Step 2, for the basic data preprocessing of the two channels, it includes the following steps: 2.

1. Perform down-conversion processing on the transmitted signal collected by the detection channel in Step 1 and the echo signal collected by the echo channel through the radio frequency front-end receiving chip at the same time sequence, and output it as a typical single-ended intermediate frequency signal. 2.

2. For the intermediate frequency signal output after the down-conversion processing by the radio frequency front-end receiving chip, use an AD high-speed analog-to-digital converter for intermediate frequency sampling to convert it into an intermediate frequency digital signal. 2.

3. The signal data output after the AD high-speed analog-to-digital conversion is divided into an I channel and a Q channel. One additional detection channel and the echo channel of the original receiver, there are a total of 4 digital signals at this time, and perform band-pass filtering on them to filter out noise interference in other frequency bands. 2.

4. After passing through the band-pass filter, combine the corresponding I channel and Q channel in the detection channel into one complex signal. Similarly, combine the corresponding I channel and Q channel in the echo channel into another complex signal.

3. The dual-channel signal quasi-coherent processing method for a magnetron radar according to claim 1, wherein: In the above Step 3, for the transmitted signal and the target echo signal after data preprocessing, connect them to the quasi-coherent correction module to obtain the coherently corrected target echo data, including the following steps: 3.

1. Perform Fourier transform on the recombined complex signal in the detection channel, and at the same time sequence, perform Fourier transform on the recombined complex signal in the echo channel. 3.

2. Take the conjugate of the digital complex signal in the detection channel after Fourier transform, and perform complex multiplication operation with the digital signal in the echo channel after Fourier transform. 3.

3. Perform inverse Fourier transform on the data after complex multiplication to obtain the coherently corrected target echo data.

4. A dual-channel quasi-coherent signal processing method based on a magnetron pulse radar according to claim 3, characterized in that: The mathematical model expression of the echo signal data output in the above Step S3.3 is as follows: Among them, E R (t) is the digital complex signal in the detection channel; E T (t) is the digital complex signal in the echo channel.

5. A dual-channel quasi-coherent signal processing method based on a magnetron pulse radar according to any one of claims 1-4, characterized in that: When performing Doppler processing in the above Step 5, first change the arrangement mode of the matrix data, then arrange them in order according to the slow time dimension, and finally pass the rearranged data through a Doppler filter, and obtain the Doppler frequency shift through spectrum analysis.

Images