RU2360360C1 - Device for linear suppression of retransmitted noise - Google Patents

Abstract

FIELD: radar-location.

SUBSTANCE: invention relates to radar-location and can be used for suppressing retransmitted noise. This technical outcome is achieved due to that, the device for linear suppression of retransmitted noise (RN) contains series connected amplitude detector, non-linear element, first multiplier, the input of which is connected to a bandpass limiter, the input of which is connected to the input of the amplitude detector. The device also contains series connected analogue-to-digital converter (ADC), first memory device, device for weighted processing, unit for every period fast Fourier transform (FFT), second memory device, noise detector, device for evaluating noise amplitude, third multiplier and a matched filter. The output of the noise detector is connected to the first input of the second multiplier, the second input of which is connected to the second output of the device for evaluating noise amplitude. The output of the second multiplier is connected to inputs of the amplitude detector and the bandpass limiter. The output of the first multiplier is connected to the first input of the third multiplier. The second input of the third multiplier is connected to the first output of the device for evaluating noise amplitude. The first output of the noise detector is connected to the input of the matched filter.

EFFECT: increased efficiency of suppressing several retransmitted signals (noise), with different delays and frequency shifting, created by an active jammer (AJ).

7 dwg

Description

The invention relates to radar and can be used to suppress relay interference.

Known devices for suppressing non-Gaussian interference presented in [1-4]. The main disadvantage of the considered devices is the low efficiency of the suppression of several simultaneously acting relay interference.

Of the known interference suppression devices, the closest in technical essence is the non-linear noise suppression (NP) device of non-Gaussian interference described in [4, p. This device contains a series-connected amplitude detector 1, a nonlinear element 2, a first multiplier 4, the second input of which is connected to the output of the band stop 3, the input of which is connected to the input of the amplitude detector 1. The output of the first multiplier 4 is the output of a nonlinear suppression device for non-Gaussian interference.

The device operates as follows. The inputs of the amplitude detector 1 and the band stop 3 receive an additive mixture of the signal reflected from the target and the relayed interference. The signal and interference are a pack of radio pulses with the same modulation. They differ in delay and / or frequency. Moreover, the signal-to-noise ratio at the input of the nonlinear suppression device is small. In the probabilistic description, the interference is non-Gaussian. The envelope of the input mixture from the output of the amplitude detector 1 is fed to the input of a nonlinear element 2, the amplitude characteristic of which is tuned to the interference parameters. In a nonlinear element, powerful non-Gaussian noise is suppressed, while a weak signal passes without significant attenuation [4]. The envelope obtained from the output of the nonlinear element 2 is fed to the input of the first multiplier 4. The phase factor from the output of the strip limiter 3 is fed to its second input. For analog signal processing, it is cos (wt + φ (t)), for digital - е ^{jφ (t)} , where φ (t) is the law of angular modulation of the input mixture. The output of the first multiplier 4 restores the law of angular modulation of the signal. In this case, the signal-to-noise ratio is close to unity. Isolation of the signal from its mixture with the remainder of the noise and combinational components can be provided by a matched filter.

The disadvantage of this device is that when exposed to several interference with different delays and frequencies, their sum forms a random process close to the Gaussian, which greatly reduces the efficiency of the nonlinear suppressor [4].

The technical result of the invention is to increase the efficiency of suppressing many relayed interference.

The essence of the invention lies in the fact that in the device for non-linear suppression of non-Gaussian interference, containing a series-connected amplitude detector, a non-linear element, a first multiplier, the second input of which is connected to the output of the band limiter, the input of which is connected to the input of the amplitude detector, additionally connected series-connected ADCs, the first storage device, weight processing device, inter-period FFT unit, second storage device, interference detector, evaluation device a interference amplitudes, a third multiplier, a matched filter, while the first input of the second multiplier is connected to the output of the interference detector, the second input of which is connected to the second output of the interference amplitude estimator, the output of the second multiplier is connected to the inputs of the amplitude detector and the band limiter, the output of the first multiplier is connected to the second input of the third multiplier, while the second output of the interference detector is connected to the input of the matched filter.

Figure 2 presents the structural diagram of the proposed device for non-linear suppression of relayed interference.

A non-linear device for suppressing relayed interference includes series-connected devices: ADC 5, first storage device 6, weight processing device 7, inter-period FFT unit 8, second storage device 9, interference detector 10, interference amplitude estimation device 11, third multiplier 13, matched filter 14 while the output of the interference detector 10 is connected to the first input of the second multiplier 12, the second input of which is connected to the second output of the device for evaluating the amplitude of the interference 11, the output of the second multiplier 12 connected to series-connected: amplitude detector 1, non-linear element 2, the first multiplier 4, the second input of which is connected to the output of the strip limiter 3, the input of which is connected to the input of the amplitude detector 1, the output of the first multiplier 4 is connected to the second input of the third multiplier 13, the second output the interference detector 10 is connected to the input of the matched filter 14, the output of the matched filter 14 is the output of the nonlinear interference suppression device.

The device is implemented in digital form and operates as follows. After digitization in the ADC 5, the analog input signal is stored in the first memory 6. Each pulse repetition period is stored as a vector of a row of samples. The accumulation time of pulses in the storage device 6 is determined by the given number of pulses K. After the accumulation time has elapsed, the sample matrix V is read from the first memory device 6 for the received burst of pulses, where the number of rows of the matrix is equal to the number of pulses in burst K, the number of columns is equal to the number of samples N used on one pulse repetition period. The weight processing device 7 performs the term-by-factor multiplication of the matrix V by the matrix W, the columns of which contain samples of the Chebyshev or Hamming window, which suppresses the side lobes of the radar image (RLI) along the frequency axis. In aperture synthesized radar (PCA), the matrix columns also contain phase correction factors for focusing the radar. All columns of the matrix W are the same (this is a sufficient condition for the possibility of first performing inter-period processing of a burst of pulses, and then performing intra-pulse processing - consistent filtering). From the output of the weighing device 7, the sample matrix is fed to the input of the block of inter-period FFT performed in columns. This operation allows you to filter out interference on the Doppler frequency, which in the future will allow you to suppress them individually. At the output of the inter-period FFT block, the matrix Y = F {V · W} is formed, where F {} is the direct FFT operation, and · is the termwise matrix multiplication operation.

The matrix Y is always complex. The view of its module in the presence of interference is shown in Fig. 3. Here n is the column number (reference number on the period), k is the row number (frequency shift number). The number of interference is 3. The input signal is set valid, its spectrum is symmetrical. Therefore, the number of interference pulses after the inter-period FFT is 6. When digitizing a quadrature signal, the number of interference pulses in FIG. 3 will be equal to the number of specified interference with different frequencies. Window processing reduces the level of the side lobes, but increases the length of the interference pulses along the frequency axis. Each of them covers up to 4-6 rows of the matrix Y. Interferences have different delays and frequencies.

The resulting matrix Y is stored in the second storage device 9, and from its output it is fed to the input of the interference detector 10. Each row of matrix Y contains samples of the received signal with a certain value of the Doppler frequency shift. This means that the samples of interference pulses with different frequency shifts lie in different rows.

. Затем отсчет огибающей сравнивается с порогом: A(k)>A_por. Если порог превышен, строка матрицы Y с номером k содержит мощную помеху и передается на первый выход обнаружителя помехи. В противном случае строка выдается на его второй выход и поступает на вход согласованного фильтра без нелинейной обработки.Interference Detector 10 analyzes the rows of matrix Y for the presence of powerful interference to be suppressed. To detect interference, threshold processing of the elements of the central column y (k) of the matrix Y is performed, k = 1 ... K. For each element of this column, the envelope value is calculated:

. Then the envelope reference is compared with the threshold: A (k)> A _por . If the threshold is exceeded, the row of matrix Y with the number k contains powerful interference and is transmitted to the first output of the interference detector. Otherwise, the line is output to its second output and enters the input of the matched filter without non-linear processing.

Detection threshold A _por = S _max · A ₂ , where S _max is the expected maximum excess of the target signal over the noise after inter-period processing; And ₂ is the median of the distribution of the array of samples of the envelope of the column A (k). To estimate the median, the samples of the array A (k) are sorted in ascending order. The median A ₂ is equal to the central reference value of the sorted array.

When choosing a threshold, it is assumed that the signal and interference are present in a small number of rows of the matrix Y (substantially less than half). Then the value of A ₂ belongs to the noise sample, is close to the median value of its distribution and depends little on the amplitude of the interference. This provides robust interference detection with different amplitudes. The factor S _max increases the threshold so that the signal from the large target could not be mistaken for interference and was not suppressed. Interferences below the threshold will also be ignored. However, such interference does not impede signal detection (it does not mask its side lobes with the uncertainty function).

,Let interference be found in the kth row of matrix Y. Then its samples are fed to the inputs of the second multiplier 12 and the device for estimating the amplitude of the interference 11. From the second output of the device for evaluating the amplitude 11, a factor is supplied to the second input of the second multiplier 12

,

where A _mid is the estimate of the amplitude of the relayed interference for the kth row of the matrix Y, A ₀ is the parameter of the amplitude characteristic (AX) of the nonlinear element, characterizing the transition of its AX through zero. The optimal AX type was determined in [4]. With a digital implementation, it is convenient to approximate this AX with a parabola, the form of which is shown in Fig. 4. In this case, A ₀ = 1.

An amplitude detector cannot be used to estimate the interference amplitude A _mid , since not all processed sample line can be covered by the interference pulse. In this case, the histogram of the distribution of the samples of the amplitudes of the kth row A (n) will contain two maxima, Fig.7. The first maximum is determined by the samples of the amplitudes of the signal and noise, not covered by an interference pulse. The second is located in the region of large amplitudes and is determined by the readings of the amplitudes of the powerful noise. It is necessary to distinguish many samples of interference under the condition of exceeding the threshold A (n)> 0.5 · A _max , where A _max is the maximum value of the amplitude in the kth row, n = 1 ... N. Estimation of interference amplitude A _mid is equal to the average value for the obtained set of samples of interference amplitudes.

The second multiplier provides the average value of the amplitude of the interference to the level A ₀ . After that, the signal and interference mixture is fed to the input of the NP device (to the input of the amplitude detector and band-pass limiter), which provides noise suppression.

. Тогда амплитудный детектор реализуется как расчет модуля:

, a полосовой ограничитель реализуется делением на огибающуюWhen implementing the NP device in digital form, processing is performed for the complex envelope of the input signal

. Then the amplitude detector is implemented as a module calculation:

, a band stop is implemented by dividing by the envelope

where φ (n) is the law of angular modulation of the input mixture.

The amplitude characteristic of a nonlinear element with A ₀ = 1 is described by the parabola g (A) = AA ² , Fig. 4. You can calculate the complex envelope of the signal at the output of the first multiplier

This allows you to simplify the digital implementation of the device NP

The expression describes the algorithm for the operation of an amplitude detector, a nonlinear element, a band stop, and the first multiplier.

подается на первый вход третьего умножителя 13, на второй вход которого подается числовой множитель k1=2·А_mid/A₀ с первого выхода устройства оценки амплитуды помехи 11. Он обеспечивает возврат уровня полезного сигнала на выходе НП к его уровню на входе. Коэффициент 2 учитывает двукратное ослабление сигнала цели в устройстве НП.From the output of the first multiplier 4 signal

fed to the first input of the third multiplier 13, the second input of which is supplied with a numerical factor k1 = 2 · A _mid / A ₀ from the first output of the interference amplitude estimator 11. It provides a return of the useful signal level at the output of the receiver to its input level. Coefficient 2 takes into account the twofold attenuation of the target signal in the NP device.

The signal from the output of the third multiplier 13 is fed to the input of a matched filter. It produces intrapulse processing and the selection of a useful signal from a mixture with residual noise and combinational components. The result of the processing is fed to the output of the device for nonlinear suppression of relayed interference.

The matched filter can be implemented by any of the known methods. With a digital implementation, it is convenient to use the fast convolution algorithm. To do this, the FFT is performed in the data line, then the resulting spectrum is multiplied by the complex conjugate spectrum of the emitted pulse, then the inverse FFT is performed.

The rows of the matrix Y arrive at the input of the matched filter in one of two ways, depending on the results of the operation of the interference detector 10. If interference is detected, the row arrives after non-linear processing from the output of the third multiplier 13. In the absence of interference, it is directly from the second output of the interference detector. This eliminates the suppression of the target signal in the absence of interference. The route of arrival of the data line is determined by the interference detector 10.

Thus, the proposed device can separately suppress several relay interference, characterized by Doppler frequency shift. At frequencies not covered by interference, a standard matched filtering of the pulse train is carried out, providing the maximum signal-to-noise ratio.

Mathematical modeling of the proposed device nonlinear suppression of relayed interference. A radar with a probing signal in the form of a coherent burst of chirp pulses was simulated. The initial data was set: the minimum frequency of the LFM pulse at the input of the ADC is 0.3 MHz, the frequency deviation is 4.4 MHz, the sampling frequency is 10 MHz; the pulse repetition period is 150 μs, the pulse duration is 37 μs, the number of pulses in a packet is 256. 512 samples were processed at each period, of which 370 correspond to the pulse duration. The number of ADC bits is 12. A Chebyshev window with a side lobe suppression depth of 100 dB was used.

Two point target signals with equal delays and different frequencies are set, as well as three interferences with different frequencies and delays. The input signal-to-noise ratio is minus 10 dB.

Figure 5 shows the cross-section of the radar range. The marks of the signal 15 and interference 16 have the same frequency. The signal-to-noise ratio at the input is minus 30 dB. Only linear processing has been performed. The signal mark exceeds the side lobes of the interference, but the margin is small. With an input signal-to-noise ratio of minus 40 dB, the signal is no longer detected.

Figure 6 shows the same cross-section of radar images after nonlinear processing with an input signal-to-noise ratio of minus 70 dB. It can be seen that the mark of signal 15 exceeds the mark of interference 16 and all of its side lobes. Mark 17 is the combinational component of the signal with interference, coinciding in frequency.

Thanks to the use of a nonlinear suppression device, the minimum detectable signal level is reduced by 40 dB.

All elements of the proposed device can be implemented in digital form, for example, on a signal processor or on a programmable logic matrix.

LITERATURE

1. US patent No. 4270223, H04B 1/12.

2. Copyright certificate 1264815, filed. 02.03.84, Н04В 1/10.

3. Copyright certificate 1311581, claimed. 11.03.85, Н04В 1/10.

4. Theory of signal detection / Ed. P.A. Bakuta. - M.: Radio and Communications, 1984.

Claims (1)

A device for non-linear suppression of relayed interference, comprising a series-connected amplitude detector, a non-linear element, a first multiplier, the second input of which is connected to the output of the band limiter, the input of which is connected to the input of the amplitude detector, characterized in that an analog-to-digital converter, the first memory device, weight processing device, inter-period fast Fourier transform (FFT) unit, second storage device oh, an interference detector, an interference amplitude estimator, a third multiplier, a matched filter, while the first input of a second multiplier is connected to the output of an interference detector, the second input of which is connected to the second output of the interference amplitude estimator, the output of the second multiplier is connected to the inputs of an amplitude detector and a bandpass limiter, the output of the first multiplier is connected to the second input of the third multiplier, the second output of the interference detector is connected to the input of the matched filter, the output of which is the output non-linear relay interference suppression devices.

Images