CN112379337A

CN112379337A - Short-distance false alarm eliminating method for long-short pulse combined pulse compression radar

Info

Publication number: CN112379337A
Application number: CN202011087871.3A
Authority: CN
Inventors: 卓智海; 杨翠翠; 蔡俊; 赵凯; 阿思哈尔·阿布尔哈斯木
Original assignee: Beijing Information Science and Technology University
Current assignee: Beijing Information Science and Technology University
Priority date: 2020-10-13
Filing date: 2020-10-13
Publication date: 2021-02-19

Abstract

The invention relates to a method for solving the problem of short-distance false alarm caused by filtering a far target in a pulse compression radar, and belongs to the field of pulse system radar signal processing. The invention finds out the interference pulse by detecting the pulse amplitude and the pulse width after the digital down-conversion, and inhibits the interference pulse, thereby solving the problem of false alarm generated in a short distance after the pulse compression. The method provided by the invention not only can solve the problem of false alarm generation in a short distance, but also is realized in an FPGA, the utilization rate of time resources is high, and the complexity of subsequent signal processing is not increased.

Description

Short-distance false alarm eliminating method for long-short pulse combined pulse compression radar

Technical Field

The invention relates to a solution of a false alarm problem caused by short distance after long pulse echo filtering of a far target in a pulse compression radar, and belongs to the field of pulse system radar signal processing.

Background

The traditional pulse system radar has irreconcilable contradiction on detection distance and distance resolution, and the pulse compression technology well solves the contradiction, so that the radar not only can have the transmitting power of long pulse, but also can obtain the distance resolution of short pulse. In order to solve the problem of short-distance shielding in searching, the early pulse system radar usually transmits a short pulse at a transmitting end first and performs signal acquisition on a short-distance section; then, a long pulse is transmitted, and signal acquisition is carried out on a long-distance segment, wherein the time sequence is shown in figure 1. Obviously, this operation is very wasteful of time resources. Therefore, the radar mostly adopts the working mode as shown in fig. 2, a long pulse is transmitted first, and then a short pulse is transmitted, and two sampling gates start to acquire signals at the same time.

After the operation mode shown in fig. 2 is adopted, in order to distinguish the long pulse from the short pulse and avoid the mutual interference between the long pulse and the short pulse, a method of modulating the long pulse and the short pulse to different frequencies and distinguishing the long pulse from the short pulse on a frequency spectrum is adopted at a transmitting end; two groups of filters are arranged at a receiving end, different mixing frequencies are generated by arranging an NCO numerically-controlled oscillator, and filtering processing is respectively carried out on the two filters, as shown in FIG. 3.

However, in practice, the filter cannot completely filter out the out-of-band pulse signals, because the pulse signals contain abundant frequencies in the process of existence or existence, and the frequencies in the pass band of the filter are necessary. Therefore, a very narrow and weak sine-like interference pulse appears in the filtered time domain, as shown in fig. 4.

When the echo signal is in the condition shown in fig. 6, the short pulse echo for detecting the short distance has left the short distance sampling gate, while the long pulse for detecting the long distance is collected by the short distance sampling gate, and digital down-conversion and pulse compression processing are performed. Since the filtering process is included in the digital down-conversion, and the frequency of the long pulse is outside the passband of the second group of filters (DDC2), a sinc-like interference pulse as shown in fig. 4 is generated, and when the pulse compression process is performed, the matching coefficient of the short pulse is matched and filtered with a very narrow pulse, and as a result, a relatively narrow pulse is generated, as shown in fig. 5, which results in a false alarm in a short range.

Disclosure of Invention

The invention provides a method for detecting and inhibiting interference pulses, which aims at the problem of false alarm caused by out-of-band interference pulses after pulse signal filtering. The method carries out time domain detection on the signals collected by the close-range sampling gate, judges whether the pulses in the signals are interference pulses or normal echo pulses, and suppresses the interference pulses, thereby solving the problem of false alarm caused by the interference pulses after filtering.

Interference pulses generated after the out-of-band pulse signal is filtered are in a sinc function shape, and besides the weak amplitude, the pulse width is also narrow, so that a basis is provided for detecting the interference pulses.

The implementation steps of the invention are summarized as follows:

the method comprises the following steps: a received signal is detected for pulses that exceed a preset amplitude threshold.

Step two: the pulse width detected in the step one does not exceed the preset pulse width threshold.

Step three: and buffering the data and recording the position of the maximum value of the pulse.

Step four: when the buffered data is output, the data near the maximum value of the pulse is set to zero.

The method provided by the invention not only can solve the problem that a remote target generates a false alarm in a short distance, but also is realized in the FPGA, the utilization rate of time resources is high, and the complexity of subsequent signal processing is not increased.

Drawings

FIG. 1: the working mode of the radar is compressed by the early pulse, so that time resources are wasted;

FIG. 2: the working mode of the existing pulse compression radar is high in time resource utilization rate;

FIG. 3: the two groups of filters separate the long and short pulses with different carrier frequencies;

FIG. 4: filtering the out-of-band pulse signal to generate a sinc function-shaped interference pulse;

FIG. 5: performing pulse compression on the filtered sinc-shaped interference pulse;

FIG. 6: an echo time sequence diagram when a short-distance false alarm occurs;

FIG. 7: leading to false alarm at close range;

FIG. 8: the method simulates the interference pulse suppression effect MATLAB;

FIG. 9: the FPGA pulse pressure result of the interference pulse is suppressed by using the method;

Detailed Description

The method is proposed to solve the practical problems encountered in engineering practice. The technical scheme of the invention is further explained by combining the drawings and the examples.

In an ideal situation, the filtering function can be well performed by strictly limiting the bandwidth of the filter. For example, in the project applied by the method, the bandwidth of the frequency modulation signal is 5MHz, the center frequency of the long pulse and the center frequency of the short pulse are different by 10MHz, and the long pulse and the short pulse can be separated by using the filter bank shown in fig. 3. In an ideal filtering result, only noise exists in the filtered signal, and no influence is generated on subsequent detection after pulse compression. When the echoes of the long pulse and the short pulse are all acquired by the same sampling gate (taking the short pulse sampling gate as an example), the residual interference pulse after the long pulse filtering can not generate any influence on the target detection due to the filtering action of the short pulse filter.

However, when the situation shown in fig. 6 occurs, the short pulse sampling gate only takes the echo signal of the long pulse, and the interference pulse remaining after filtering will have a major influence on the pulse compression, and the input signal in fig. 5 is the filtered interference pulse. Fig. 7 shows the FPGA pulse compression result in the case of fig. 6, where the first 210 is the short-range pulse compression result and the last 770 is the long-range pulse compression result, where the red part is the false-alarm pulse generated in the case of fig. 6.

The present invention focuses on how to detect and suppress the interference pulse in the short-distance segment in response to the false alarm condition. The specific implementation process is as follows:

firstly, setting a signal amplitude threshold A and a pulse width threshold K, and detecting the amplitude and the pulse width of the filtered signal. When the signal exceeds an amplitude threshold A, the signal is considered to be a non-noise signal; when the pulse width exceeds the width threshold K, a normal echo signal is considered. This is because the interference pulses are very narrow, and therefore the pulse width is taken as the main distinguishing feature between the interference pulses and the normal echo pulses. Here, a is tentatively 100; let K be 1.5us temporarily, and the corresponding data point number be 15.

And after the effective data output by the filter (II) arrives, writing the data into FIFO for buffering. And detecting the amplitude and the width of the data written into the FIFO, and recording and buffering the position of the peak value of the interference pulse. After all the data are input into the FIFO, the data in the FIFO are derived, and when the data are derived, the data at M points on the left and right of the peak position of the interference pulse are set to zero, and the provisional M is 12.

The detection and zeroing of the interference pulses will be described in detail below.

And (III) detecting the signal amplitude. Firstly, judging whether the data has a value exceeding a threshold amplitude A, if not, considering that the input data is all noise, and pulling down the false alarm flag bit. Here a is chosen to be above the system noise but not so much above that it would affect the detection of the interference pulses. Data after digital down conversion in the FPGA is divided into an I path and a Q path which are orthogonal, the peak of interference pulse may be negative in the I path and positive in the Q path or just opposite, and therefore, a judgment condition can be set during detection to be that the data of the I path or the Q path exceeds a threshold A. The same applies to the detection of the peak point position.

And (IV) recording the position of the peak point. And (3) judging whether the data of the current position is the peak value peak or not on the basis of the step (three). Because the interference pulse is a sinc-shaped function curve, the part exceeding the amplitude threshold A is increased firstly and then reduced, and only one extreme point exists. So if the current data is larger than the surrounding data, its position is recorded and buffered, here indicated by the signal index. Since the FIFO has a counter both when writing and when reading data, the data location can be marked by the counter.

And (V) detecting the pulse width. And (3) judging the pulse width exceeding the threshold K on the basis of the third step. The width of the pulse and the number of data points are in one-to-one correspondence, so the pulse width is judged by detecting the number of points. A counter id _ cnt is set, which is incremented by 1 each time a point exceeds the threshold amplitude a. And (5) when one PRT is finished, if the value of id _ cnt does not exceed the pulse width threshold K, considering that an interference pulse exists and a false alarm exists, pulling the false alarm flag high, and continuing to execute the step (six). Otherwise, if the value of id _ cnt exceeds K, the signal is considered to be a normal echo pulse, and at this time, the interference pulse does not affect the signal detection, and the false alarm flag is pulled down, so that the FIFO outputs data normally.

And (VI) removing the interference pulse. On the basis of (five), the FIFO counts when outputting data, and the counter is rd _ cnt. The peak position index has been recorded In (IV), and when the value of rd _ cnt is between (index-M, index + M), the FIFO output data is replaced with 0, thereby removing the glitch. The Index may have multiple values, indicating that there are multiple glitches, and is buffered when the FIFO inputs data. And updating the value of the index once when the FIFO starts to output data, and updating the value of the index once every time the value of rd _ cnt reaches index + M +1 until the FIFO outputs all data and the index is cleared.

Because the number of short pulse points is small, the time sequence often has enough margin, and therefore, through the time sequence verification, the method does not increase the tension on the time sequence and the complexity of subsequent signal processing.

FIG. 8 is a MATLAB simulation of the present method, wherein the red curve is the result after using the present method. It can be seen that an interference pulse is generated after filtering the original pulse signal, and a section of false alarm pulse exists after pulse pressure; after the treatment by the method, the interference pulse is suppressed, and the false alarm after the pulse pressure is relieved.

FIG. 9 shows the FPGA test result of the present method. In comparison with fig. 7, it can be seen that the false alarm pulse after the short-range pulse pressure has been eliminated after the method is used.

Claims

1. A short-distance false alarm eliminating method for long and short pulse combined pulse compression radar is characterized in that: the radar with a pulse compression system of a long-short pulse mode is adopted to detect signals in a short-distance range by setting an amplitude threshold and a pulse width threshold, and suppress narrow interference pulses after filtering.

2. The method for solving the short-range false alarm of the pulse compression radar as claimed in claim 1, characterized by the following steps:

3. The method for solving the short-range false alarm of the pulse compression radar as claimed in claim 1, wherein: the long pulse and the short pulse are not in the same frequency point and can be separated by a filter.