A method for detecting targets in the short-range blind area based on piecewise clutter map constant false alarm

Abstract

The present invention relates to radar signal processing, and particularly to a method for detecting targets in the short-range blind area based on piecewise clutter map constant false alarm. The present invention comprehensively utilizes technologies such as piecewise constant false alarm, full-channel clutter map, and selecting the larger one from the clutter map constant false alarm backgrounds to complete target detection in the short-range blind area using partial pulses, while ensuring that the processing gain basically does not decrease, which can support the radar to complete target detection in the short-range blind area without additionally requiring time resources for target detection in the short-range blind area.

Description

Technical Field

[0001] The present invention relates to a method for detecting targets in a short - range blind area based on piece - wise clutter map constant false alarm, and detects targets in a short - range blind area based on methods such as piece - wise constant false alarm, full - channel clutter map, and selecting the larger one from clutter map constant false alarm backgrounds. Background Art

[0002] In modern warfare, the role of strategic early warning has become increasingly prominent. It is often required that for the main surveillance direction, as soon as the enemy's combat platform takes off, our radar can detect and track it in real time. Therefore, the requirement for the detection range of the radar is getting higher and higher. To increase the detection range, the radar often uses technologies such as large time - width pulses to solve the problem. Although this technology can effectively improve the radar detection range, it also brings problems such as an overly large short - range blind area, making it impossible for the radar to effectively take into account the training and support of our own aircraft near the position. Currently, to solve the short - range blind area problem, there are generally the following methods:

[0003] 1) Short - range blind - area compensation pulse

[0004] Before the large time - width pulse, a small time - width pulse is specially designed, usually in the order of dozens of microseconds in length. At the same time, a period of silent reception is designed, usually in the order of hundreds of microseconds (equal to the time width of the large time - width pulse), for detecting targets in the short - range blind area. Although this method can effectively solve the short - range blind area problem, it also brings new problems. It requires reserving dedicated time resources for detecting targets in the short - range blind area, reducing the utilization rate of the radar's time resources, reducing the available time of the radar's large time - width pulse, resulting in fewer accumulable numbers of the large time - width pulse, a decrease in the accumulated power, and a decrease in the maximum detection range of the radar.

[0005] 2) Detecting targets using the half - pulse compression area

[0006] The reason why the large time - width pulse cannot detect short - range targets is that the radar cannot receive signals during the energy emission period. Therefore, the wider the pulse, the larger the short - range blind area. Some radars, in order not to sacrifice time resources as much as possible, basically do not design dedicated blind - area compensation pulses, but adopt a strategy of detecting targets after receiving part of the energy, that is, a technology of detecting targets using the half - pulse compression area. In the short - range blind area, since the radar cannot receive during the emission period, the target echo pulses received in the short - range blind area are incomplete, and only part of the pulses are received. The closer the target distance, the fewer the remaining target echo pulses that can be received. When using this part of the remaining pulses for target detection, due to the incomplete pulses, the processing gain drops rapidly, resulting in a rapid drop in the target detection probability and impaired radar short - range detection ability. Summary of the Invention

[0007] In view of the deficiencies of the background art, the present invention comprehensively utilizes techniques such as segmented constant false alarm, all-channel clutter map, and selecting the larger one from the clutter map constant false alarm backgrounds, etc. It uses some pulses to complete target detection in the short-range blind area, while ensuring that the processing gain basically does not decrease, which can support the radar to complete target detection in the short-range blind area without additionally requiring time resources for target detection in the short-range blind area.

[0008] The technical solution of the present invention is: A method for target detection in the short-range blind area based on segmented clutter map constant false alarm, including the following steps:

[0009] Step 1: In view of the incomplete characteristics of the target echo pulses in the short-range blind area, a special segmented constant false alarm is adopted to divide the short-range blind area into multiple segments. The fewer the remaining echo pulses, the more guard cells are used when estimating the constant false alarm background.

[0010] Step 2: Perform all-channel clutter map estimation. Different from general radars that only perform clutter estimation in the ground clutter channel, the present invention performs clutter map estimation in all processing channels of the radar.

[0011] Step 3: For the constant false alarm estimation and clutter map estimation of the same processing channel of the same data unit, select the larger one as the background estimation.

[0012] Step 4: Use an independent threshold for decision-making. This threshold is different from the threshold of the radar in the non-short-range blind area. Those exceeding the threshold are recognized as targets.

[0013] The advantages of the present invention are as follows: The method of the present invention can significantly improve the target detection ability of the radar in the short-range blind area without additional consumption of time resources. Traditional methods still use the classic constant false alarm processing method in the short-range blind area, without considering that the incomplete pulses of the target echo will lead to incomplete pulse compression (hereinafter referred to as half pulse compression), waveform broadening after pulse compression, and signal-to-noise ratio reduction. This method formulates a special processing method according to the characteristics of half pulse compression, associates the number of selected guard cells with the remaining ratio of the pulses that can receive the target echo, and ensures that the constant false alarm background estimation is not artificially overestimated due to waveform broadening after pulse compression, resulting in a decrease in signal-to-noise ratio. Different from traditional methods that only perform clutter map processing in the ground clutter channel, the present invention uses all-channel clutter map processing, selects the larger one from the clutter map background estimation and the constant false alarm background estimation, and further reduces the false alarm probability. In order to further improve the target detection ability of the radar in the short-range blind area, an independent threshold is used for decision-making. Description of the Drawings

[0014] Figure 1 Comparison results of full pulse compression and half pulse compression.

[0015] Figure 2 It is a flow block diagram of a method for target detection in the short-range blind area based on segmented clutter map constant false alarm. Detailed implementation manners

[0016] Glossary: Pulse width: The duration of the radar pulse.

[0017] PRI: Radar repetition interval.

[0018] Half pulse compression region: In a pulsed radar system, during the transmission of a pulse, reception is not possible. As a result, the target echo received during the transmission duration is incomplete, leading to waveform broadening after pulse compression and a decrease in processing gain. This region is called the half pulse compression region.

[0019] Protection unit: To prevent inaccurate background estimation of constant false alarm rate (CFAR), several directly adjacent units on the left and right of the current unit are removed and do not participate in the background estimation of CFAR.

[0020] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0021] As Figure 1 shown, the method of the present invention includes the following steps. Step 1: Divide the near - range blind zone into 7 segments, and determine the number of protection units for CFAR background estimation according to the remaining proportion of pulses that can receive target echoes. The specific remaining proportions of pulses for the 7 segments are shown in the following table:

[0022] Table 1 Protection unit values

[0023] Remaining ratio (a) Protection unit (b) a≥80％ b＝2 70％≤a＜80％ b＝4 60％≤a＜70％ b＝5 50％≤a＜60％ b＝6 40％≤a＜50％ b＝7 30％≤a＜40％ b＝8 a＜30％ b＝10

[0024] Table 1 shows the best parameter ratio based on engineering practice, which can improve the detection ability of the near - range blind distance.

[0025] Perform CFAR background estimation on the echo data of each PRI. Generally, when the radar performs CFAR background estimation, 2 protection units are usually taken, which is feasible when the target echo pulse is completely received. However, in the half pulse compression region, since the received target echo pulse is incomplete, non - matched filtering will occur during pulse compression, and the result after pulse compression will be broadened, as shown in the following figure. The original pulse width was 300 microseconds, and full pulse compression of 300 - microsecond echo (right), half pulse compression of 200 - microsecond echo (middle), and half pulse compression of 100 - microsecond echo (left) were performed respectively. It can be seen that as the pulse width that can be processed continuously decreases, the result after pulse compression becomes continuously broadened. Using the conventional 2 protection units to calculate the CFAR background estimation will make the CFAR background estimation at the target position higher, resulting in the target not being detected.

[0026] The constant false alarm background estimation adopts the unit average selection method. The radar generally has 10 to 20 MTD processing channels, of which the first processing channel is generally used to process ground clutter. A radar scan generally has 3000-5000 PRIs, which can be equated to the direction. Usually a PRI has 4000-6000 data units, which are arranged in order of distance. The smaller the number, the closer the distance. Take the data units of the close-range blind area, usually 600-800 units. Assume Represents the constant false alarm background estimate of the jth data unit of the i-th PRI of the k-th processing channel. N i It is calculated by the following formula:

[0027]

[0028] In the formula, represents the jth PRI of the i-th processing channel 1 The amplitude value of the data unit is generally taken from the jth data unit of the ith PRI of the kth processing channel, and the left and right t data units are generally taken. The left and right b protection units need to be skipped and do not participate in the estimation to ensure the correct noise estimation. The value of b is determined according to the remaining proportion of the echo pulse that can be received by the current data unit with reference to Table 1. The mean of the selected left and right t units is calculated respectively (t is generally 8), and the mean 1 (right) and the mean 2 (left) are obtained. After averaging, the constant false alarm background estimation of the current data unit in the kth processing channel is obtained. The constant false alarm background estimation is performed on the data unit of each processing channel. Through the steps, the signal-to-noise ratio can be improved, thereby improving the detection capability of the target signal.

[0029] Step 2: Take the radar data after MTD processing, build a clutter map for each processing channel, and perform clutter map background estimation on all channels. The clutter map estimation uses the following formula.

[0030]

[0031] In the formula, is the clutter estimate of the kth processing channel when the radar last scanned this unit (the jth data unit of the i-th PRI), is the clutter update estimate of the kth processing channel of this unit (the jth data unit of the i-th PRI) during the current radar scan, It is the clutter estimate of the kth processing channel of the current scanned unit (the jth data unit of the i-th PRI). α is the coefficient for adjusting the update speed. The smaller the coefficient, the faster the update. Generally, α=7 / 8. The clutter map is updated once every time the radar scans a circle. As time goes by, the estimate of each clutter unit is recursively updated, so after a period of time, the clutter map stores the background level of the corresponding clutter unit. It is determined by the following formula.

[0032]

[0033] In the formula, represents the amplitude value of the j-th data unit of the i-th PRI of the k-th processing channel. When estimating the clutter map, with the current unit (azimuth i, range j) as the central unit, the average value of the amplitude values of all units in the area with an azimuth span of (2c + 1) and a range span of (2d + 1) is calculated as The values of c and d are related to the parameters of the radar itself. c is generally determined according to the radar beam width and usually takes values between 10 and 20. d is related to the radar range resolution and the average size of clutter and usually takes values between 8 and 10.

[0034] Step 3: Perform a maximum selection on the obtained in Step 1 and the obtained in Step 2. Compare the data. Take the same data units of the same PRI in the same processing channel for comparison, and select the larger of the two estimations as the estimation of the current unit. Since the near range is a clutter environment, by comparing with the clutter map, the environmental adaptability of this method can be improved.

[0035] Step 4: Make a decision using an independent threshold. Generally, radars use the same threshold for decision-making, that is, regardless of the range, a single threshold is used for processing. Considering that the benefit of processing the near-range blind area is insufficient, this invention divides this area into 4 segments, and uses an independent threshold in this area. The threshold is determined according to the remaining pulse ratio of the receivable target echo. The specific division is shown in the following table.

[0036] Table 2 Threshold Values

[0037] Remaining ratio (a) Threshold (th) a≥80％ th = 10 dB 50％≤a＜80％ th = 9 dB 30％≤a＜50％ th = 8 dB a＜30％ th = 7 dB

[0038] Table 2 shows the best parameter ratio according to engineering practice.

[0039] The threshold decision is shown by the following formula.

[0040]

[0041] In the formula, represents the amplitude value of the j-th data unit of the i-th PRI of the k-th processing channel. represents the constant false alarm background estimation of the j-th data unit of the i-th PRI of the k-th processing channel. is the clutter estimation of the k-th processing channel for the current scan to this unit (the j-th data unit of the i-th PRI). th is determined according to the remaining pulse ratio of the receivable target echo.

[0042] After the above calculation, if CA > 0, it is determined that the current point crosses the threshold and can be further processed; otherwise, it is determined that the current point does not cross the threshold.

Claims

1. A method for detecting targets in the short - range blind area based on piece - wise clutter map CFAR, characterized in that: It includes the following steps: Step 1, adopt piece - wise CFAR and divide the short - range blind area into multiple segments; Specifically: Divide the short - range blind area into 7 segments, and determine the value of the guard cell b according to the pulse remaining ratio value a of the received target echo; The guard cell is to remove several directly adjacent cells on the left and right of the current cell to prevent inaccurate clutter map background estimation and does not participate in the clutter map background estimation; Let represent the constant false alarm background estimate of the j-th data unit of the i-th PRI of the k-th processing channel, then: In the formula, represents the amplitude value of the j 1 -th data unit of the i-th PRI of the k-th processing channel, and t represents the number of selected left and right units; Step 2, perform clutter map estimation for all channels. The clutter map estimation formula is: Wherein, is the clutter estimation of the k-th processing channel when the previous radar scans this unit, and α is the coefficient for adjusting the update speed, is the updated clutter estimation of the k-th processing channel of this unit during the current radar scan, is the clutter estimation of the k-th processing channel when this unit is scanned currently; Among them is as follows: Wherein, represents the amplitude value of the j-th data unit of the i-th PRI of the k-th processing channel, c is determined according to the radar beam width, and the value of d is related to the radar range resolution and the average size of clutter; Step 3, for the CFAR estimation and clutter map estimation of the same processing channel of the same data unit, select the larger value as the background estimation; Step 4, make a decision using different thresholds according to the pulse remaining ratio value of the received target echo.

2. The method for detecting targets in the short - range blind area based on piece - wise clutter map CFAR according to claim 1, characterized in that: In step 4, the specific process of making a decision using different thresholds is as follows: The relationship between the remaining ratio and the threshold is shown in the following table: The threshold decision formula is: In the formula, represents the amplitude value of the j-th data unit of the i-th PRI of the k-th processing channel; represents the constant false alarm background estimation value of the j-th data unit of the i-th PRI of the k-th processing channel; is the clutter estimation value of the k-th processing channel when scanning to this unit currently; th is determined according to the pulse remaining ratio of the receivable target echo; if CA > 0, it is judged that the current point crosses the threshold, otherwise it is judged that the current point does not cross the threshold.

3. The method for detecting targets in the short - range blind area based on piece - wise clutter map CFAR according to claim 1, characterized in that: The specific values of the remaining ratio and the guard cell are:

4. The method for detecting targets in the short - range blind area based on piece - wise clutter map CFAR according to claim 1 or 2, characterized in that: t takes the value of 8.

5. The method for detecting targets in the short - range blind area based on piece - wise clutter map CFAR according to claim 1 or 2, characterized in that: The coefficient α of the update speed takes the value of 7 / 8.

6. The method for detecting targets in the short - range blind area based on piece - wise clutter map CFAR according to claim 1 or 2, characterized in that: c takes values between 10 and 20.

7. The method for detecting targets in the short - range blind area based on piece - wise clutter map CFAR according to claim 1 or 2, characterized in that: d takes values between 8 and 10.

Images