Channel fixed time delay error estimation method for wideband digital array radar

Abstract

The application provides a channel fixed time delay error estimation method of a wideband digital array radar, comprising: collecting and storing each channel output signal containing a channel fixed time delay, and performing parallel-serial conversion on the stored signal; inputting each channel signal after parallel-serial conversion into a pulse compression module in time, performing peak value detection on a pulse compression output signal, and recording a peak value point and sampling time points corresponding to the peak value point and two points before and after the peak value point; using the peak value point and the two points before and after the peak value point to fit a polynomial with the peak value point as a reference zero point; adding an integer part of the channel fixed time delay and a fractional part of the channel fixed time delay to obtain a fixed time delay of each channel, and then calculating a fixed time delay error between each channel and a reference channel. The application can achieve an estimation accuracy of 0.001 to 0.014 times of a sampling interval under the condition that a signal-to-noise ratio of an input signal is 10 dB to 30 dB, can reduce a hardware resource occupancy rate, can improve a performance of digital beam forming in the wideband digital array radar, and can improve a system performance of the wideband digital array radar.

Description

TECHNICAL FIELD

[0001] The application discloses a channel fixed time delay error estimation method of a wideband digital array radar. BACKGROUND

[0002] As a key technology of wideband digital array radar, the performance of wideband digital beam forming influences the ability of target detection, tracking and identification of the radar system. However, for a large-scale wideband digital array radar, the array channels are usually hundreds of channels, and the signal asynchronization caused by the system fixed time delay error of each channel will inevitably cause the performance decline of the wideband digital beam forming. Meanwhile, due to the large number of array channels, the system requires more hardware resources. Therefore, the synchronization of high-speed digital signals between channels and the large demand of system hardware resources have been the main technical problems hindering the engineering application of the wideband digital beam forming technology.

[0003] In order to solve the high-speed signal synchronization problem caused by the channel fixed time delay error, the time delay estimation algorithm in array signal processing can be used to estimate the channel fixed time delay, and the time delay compensation is performed on the signals of each channel to achieve the signal synchronization. At present, the time delay error estimation research applied to the radar array channels is less, and most of the researches on time delay estimation are concentrated in the field of underwater acoustic array signal processing. For example, the variable step size algorithm based on iteration time, the time delay estimation algorithm based on fractional Fourier transform, and the introduction of neural network, machine learning and genetic algorithm into the fractional time delay estimation algorithm. However, due to the high algorithm complexity, the application field is also limited. Therefore, the selection of time delay estimation algorithm is different in different fields, and the time delay estimation accuracy and algorithm complexity need to be balanced according to the actual system requirements. SUMMARY

[0004] The technical problem solved by the application is to overcome the shortcomings of the prior art, and to provide a channel fixed time delay error estimation method of a wideband digital array radar, which can achieve fractional time delay estimation accuracy, can effectively overcome the influence of the channel fixed time delay error on the wideband digital beam forming, reduce the hardware resource occupancy rate, and improve the performance of the beam forming in the wideband digital array radar system.

[0005] The technical solution of the application is a channel fixed time delay error estimation method of a wideband digital array radar, which comprises the following steps:

[0006] Step 1: after a same hour-wide linear frequency modulation signal is fed into each array receiving channel of the wideband digital array radar, the output signals of each channel containing the channel fixed time delay are collected and stored, and the stored signals are converted into parallel-serial conversion and then output to a pulse compression module;

[0007] Step two, the parallel-serial converted channel signals are input into the pulse compression module, the pulse compression output signal is input into the peak detection module, the peak value point and the corresponding sampling time of the two points before and after the peak value point are recorded, and the sampling time of the peak value point can be considered as the integer part of the channel fixed time delay;

[0008] Step three, taking the peak value point as the reference zero point, a polynomial is fitted using the peak value point and the two points before and after the peak value point, and the abscissa of the polynomial peak value point is recorded, which is the fractional part of the channel fixed time delay;

[0009] Step four, the integer part and the fractional part of the channel fixed time delay are added to obtain the fixed time delay error between each channel and the reference channel, which can be accurately compensated by using a time domain fractional delay filter.

[0010] In the above channel fixed time delay error estimation method of a wideband digital array radar, in step two, the sampling time of the pulse compression module output peak value point can be considered as the integer part of the channel fixed time delay, and the specific calculation method is as follows:

[0011] S1, the fixed time delay estimation of the array channel can be regarded as the estimation of the unknown time delay parameter t0 of the time delay signal s(t-t0) under the Gaussian white noise n(t) with power spectral density N0 / 2, that is, the estimation of the unknown time delay parameter t0 in the following formula:

[0012] y(t) = s(t-t0) + n(t) 0 < t < T p ,

[0013] Where, T p is the pulse width.

[0014] S2, the maximum likelihood estimation of the likelihood function of y(t) in S1 is solved, and the maximum likelihood estimation of the time delay t0 is obtained, that is, the correlation integral of the partial derivative of the received signal y(t) and the time delay signal s(t-t0) with respect to t0 is 0, as shown in the following formula:

[0015]

[0016] S3, the formula in S2 is further simplified by using the total differential formula, and the formula is as follows, so the maximum likelihood estimation in S2 is the time when the correlation integral of the signal y(t) and s(t-t0) is maximum, that is, the time when the correlation output peak value of the signal y(t) and s(t-t0) is, that is, the time t int corresponding to the pulse compression peak value.

[0017]

[0018] Where, t intThe integer part of the channel fixed time delay estimation can be considered.

[0019] In the channel fixed time delay error estimation method of the wideband digital array radar, in step three, the specific calculation method of the fractional channel fixed time delay is:

[0020] S4, the values REG7, REG8 and REG9 output by the peak value estimation module are recorded as (0, f(0)), (-1, f(-1)) and (1, f(1)) respectively;

[0021] S5, a quadratic polynomial as f(t) = at 2 +bt+c is fitted according to the above three points, and the abscissa of the peak value is calculated according to the following formula:

[0022]

[0023] Wherein, t decimal is the fractional part of the channel fixed time delay estimation.

[0024] In the channel fixed time delay error estimation method of the wideband digital array radar, in step three, the channel fixed time delay error can be calculated according to the following formula:

[0025]

[0026] Wherein, Δτ i is the estimated channel fixed time delay error.

[0027] The channel fixed time delay error estimation method of the wideband digital array radar comprises the following steps: step one, after the same linear frequency modulation signal is fed into each array receiving channel of the wideband digital array radar, the channel output signals containing the channel fixed time delay are collected and stored, and the stored signals are converted into parallel-serial conversion and output to the next module; step two, the parallel-serial converted channel signals are input into the pulse compression module in time, the peak value of the pulse compression output signal is detected, and the sampling time and the peak value corresponding to the peak point and the two points before and after the peak point are recorded, and the sampling time of the peak point can be considered as the integer part of the channel fixed time delay; step three, taking the peak point as the reference zero point, a polynomial is fitted using the peak point and the two points before and after the peak point, and the abscissa of the peak point of the polynomial is recorded, which is the fractional part of the channel fixed time delay; step four, after the integer part and the fractional part of the channel fixed time delay are added, the fixed time delay error between each channel and the reference channel is calculated. The estimation accuracy of the present application reaches 0.001 to 0.014 times of the sampling interval under the condition that the signal-to-noise ratio of the input signal is 10dB to 30dB, the hardware resource occupation rate can be reduced, the performance of the digital beam forming in the wideband digital array radar can be improved, and the system performance of the wideband digital array radar is improved.

[0028] The beneficial effects of the present invention compared with the prior art are:

[0029] (1) The present invention introduces the delay estimation positioning algorithm in the underwater acoustic array into the channel delay estimation error estimation of the digital array radar. Compared with integer multiple delay estimation methods such as the correlation method, the estimation performance of the channel delay error can be significantly improved.

[0030] (2) Compared with algorithms such as variable step size based on iteration time and delay estimation based on fractional Fourier transform, the present invention fully utilizes the pulse compression link in the wideband digital array radar, does not need to add additional processing links in the integer delay estimation part, and can reduce hardware resource usage.

[0031] (3) The present invention has a simple design, small amount of calculation, and is easy to implement in engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of a method for estimating fixed time delay errors of channels of a wideband digital array radar according to the present invention;

[0033] Figure 2 This is an example diagram of peak detection in the present invention;

[0034] Figure 3 This is an example diagram of interpolation of fractional delay estimation in the present invention;

[0035] Figure 4 This is an example diagram of indicators related to delay estimation accuracy under different signal-to-noise ratios of the present invention;

[0036] Figure 5 This figure illustrates the impact of delay error on beamforming and the performance improvement of the beamforming pattern after delay estimation compensation in the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described below in conjunction with the embodiments.

[0038] like Figure 1 As shown, the present invention provides a channel fixed delay error estimation method for a wideband digital array radar, comprising the following steps:

[0039] Step 1: After the same hour-width linear frequency modulation signal is fed into each array receiving channel of the broadband digital array radar, the output signal of each channel with a fixed channel delay is collected and stored, and the stored signal is converted from parallel to serial and then output to the pulse compression module;

[0040] Step two, the parallel-serial converted channel signals are input into the pulse compression module, the pulse compression output signal is input into the peak detection module, the peak value point and the corresponding sampling time of the two points before and after the peak value point are recorded, and the sampling time of the peak value point can be considered as the integer part of the channel fixed delay;

[0041] Figure 2 The schematic diagram of peak detection is given, and the specific method can be represented as:

[0042] S1, after the pulse compression signal is input, the counter is started, and the sizes of the input signals of the three adjacent time points are recorded as REG1, REG2 and REG3;

[0043] S2, the sizes of the values REG1, REG2 and REG3 are compared, if there is a value REG2 greater than the values REG1 and REG3, the values REG1, REG2 and REG3 are saved, and the corresponding values are REG4, REG5 and REG6;

[0044] S3, the size of REG5 and REG8 is compared, if the value REG5 is greater than the value REG8, the values REG4, REG5 and REG6 are saved, the corresponding values are REG7, REG8 and REG9, and the counting value of the counter is saved;

[0045] Repeat S7 to S9 until the pulse compression signal is input. At this time, the values of REG7, REG8 and REG9 are the pulse compression peak value and the values of the two points adjacent to the peak value, and at the same time, the counting value of the counter can be considered as the integer part of the channel fixed delay.

[0046] Step three, taking the peak value point as the reference zero point, a polynomial is fitted using the peak value point and the two points before and after the peak value point, and the abscissa of the peak value point of the polynomial is recorded, which is the fractional part of the channel fixed delay;

[0047] Figure 3 The interpolation example diagram of the fractional delay estimation is given, and the specific method can be represented as:

[0048] S4, the values REG7, REG8 and REG9 output by the peak value estimation module are recorded as (0, f(0)), (-1, f(-1)) and (1, f(1)) respectively;

[0049] S5, a quadratic polynomial as f(t) = at 2 +bt+c is fitted according to the above three points, and the abscissa of the peak value is calculated according to the following formula:

[0050]

[0051] Where, t decimalFractional part of the channel fixed delay estimation.

[0052] Step four, add the integer part and the fractional part of the channel fixed delay to obtain the fixed delay error between each channel and the reference channel, which can be compensated precisely by the time-domain fractional delay filter.

[0053] Figure 4 The average value, standard deviation and root mean square error of the channel fixed delay error estimation algorithm under different signal-to-noise ratios are given by 1000 times of Monte Carlo simulation experiments.

[0054] It can be obtained from Figure 4 that under the condition of signal-to-noise ratio of 10dB-30dB, the standard deviation of the channel fixed delay estimation is only 0.011-0.003 times of the sampling interval, the dispersion degree of the obtained estimation value is low, and the stability of the estimation algorithm is good; at the same time, the mean square error of the delay estimation is only 0.012-0.001 times of the sampling interval, the mean value is 0.009-0.001 times of the sampling interval, and the error of each channel after estimation and compensation is only in the order of picoseconds.

[0055] Figure 5 The generalized directional pattern of digital beam forming before and after estimation and compensation of channel fixed delay error is given.

[0056] It can be obtained from Figure 5 that when the standard deviation of the channel fixed delay error is 20ps, compared with the case without channel fixed delay error, the beam pointing is shifted by 0.6°, the normalized main lobe gain is reduced by 1.060dB, the normalized peak side lobe and average side lobe level are increased by 10.349dB and 6.437dB respectively, and after compensation of the channel fixed delay error, the beam pointing and the normalized main lobe gain do not change significantly, only the peak side lobe and the average side lobe level increase slightly.

[0057] Although the content of the present application has been described in detail by the above preferred embodiments, it should be recognized that the above description should not be considered as a limitation of the present application. After reading the above content, various modifications and alternatives of the present application will be obvious to those skilled in the art. Therefore, the protection scope of the present application should be defined by the appended claims.

Claims

1. A method for estimating channel fixed time delay error of a wideband digital array radar, characterized in that, The application relates to a method for estimating fixed time delay of each channel of a wideband digital array radar. Step one: after the same hour-wide linear frequency modulation signal is fed into each array receiving channel of a wideband digital array radar, each channel output signal containing a channel fixed time delay is collected and stored, and the stored each channel output signal is parallel-serial converted and output to a pulse compression module; Step two: each channel output signal after parallel-serial conversion is input into the pulse compression module in time, pulse compression output signals are obtained, the pulse compression output signals are input into a peak value detection module, and the peak value point, the sampling time of the peak value point and the peak values of the two points before and after the peak value point are recorded; the sampling time of the peak value point is the integer part of the channel fixed time delay; Step three: taking the peak value point as a reference zero point, a polynomial is fitted by using the peak value point and the two points before and after the peak value point, and the abscissa of the polynomial peak value point is recorded as the fractional part of the channel fixed time delay; Step four: after the integer part of the channel fixed time delay and the fractional part of the channel fixed time delay are added, the fixed time delay error between each channel fixed time delay and a reference channel is calculated.

2. The channel fixed time delay error estimation method for a wideband digital array radar according to claim 1, characterized in that, In the step one, The in-phase or quadrature data sampling frequency obtained after intermediate frequency sampling and digital down conversion is 4 times the frequency of the FPGA main clock, and the in-phase or quadrature signal of each channel is input into a framing and parallel-serial conversion module after 4 times extraction.

3. The channel fixed time delay error estimation method for a wideband digital array radar according to claim 2, characterized in that, In the step one, The bit width of the input data of the parallel-serial conversion module is 112 bits, and every 28 bits from low to high is the in-phase or quadrature signal obtained after digital down conversion, wherein the low 14 bits are the in-phase signal and the high 14 bits are the quadrature signal; the each channel signal after parallel-serial conversion is sequentially output according to the channel number order, so that the in-phase and quadrature signals of one sampling of each channel are output in each clock cycle, and the time-sharing processing is realized.

4. The channel fixed time delay error estimation method for a wideband digital array radar according to claim 3, characterized in that, In the step two, The sampling time of the peak value point is the integer part of the channel fixed time delay, and the method comprises the following steps: S1, the fixed time delay estimation of the array channel is regarded as the estimation of the unknown time delay parameter of the time delay signal s (t-t0) under the Gaussian white noise n (t) with the power spectral density N0 / 2, that is, the estimation of the unknown time delay parameter t0 in the following formula; y(t) = s(t - t0) + n(t) 0 < t < T p , wherein T p is the pulse width; S2, the maximum likelihood estimation of S1 is solved by making the likelihood function of y (t) maximum, and the maximum likelihood estimation of the time delay t0 is that the correlation integral of the partial derivative of the received signal y (t) and the time delay signal s (t-t0) with respect to t0 is 0, as shown in the following formula: S3, using the total differential formula to S2 in the formula to further simplify, the formula is shown as follows, so the maximum likelihood estimator in S2 is the time corresponding to the maximum value of the correlation integral of signal y(t) and s(t-t0), that is, the time corresponding to the peak value of the correlation output of signal y(t) and s(t-t0), that is, the time t corresponding to the pulse compression peak int ; where t int is the integer part of the channel fixed delay estimate.

5. The channel fixed time delay error estimation method for a wideband digital array radar according to claim 4, characterized in that, In the step two, the pulse compression module comprises a signal clipping, a matched filter and a coordinate rotation digital calculation part.

6. The channel fixed time delay error estimation method for a wideband digital array radar according to claim 5, characterized in that, In the step two, the pulse compression module is used for: S4, the 28-bit in-phase and quadrature signals are clipped to obtain 14-bit in-phase and quadrature signals; S5, the in-phase and quadrature signals are respectively input into FIR filters to calculate the time domain convolution of the in-phase and quadrature signals and the real part and the imaginary part of the baseband transmitting signal; S6, the real part and the imaginary part obtained are added respectively, and then input into a coordinate rotation digital module to calculate the modulus value of the complex number, so that the pulse compressed output signal is obtained.

7. The channel fixed time delay error estimation method for a wideband digital array radar according to claim 6, characterized in that, In the step two, the peak value detection module is used for: S7, after the input of the pulse compression signal, start the counter and record the size of the input signal of the adjacent three time points as REG1, REG2 and REG3; S8, compare the size of the values REG1, REG2 and REG3, if there is a value REG2 greater than the values REG1 and REG3, save the values REG1, REG2 and REG3, and the corresponding values are REG4, REG5 and REG6 respectively; S9, compare the size of REG5 and REG8, if the value REG5 is greater than the value REG8, save the values REG4, REG5 and REG6, and the corresponding values are REG7, REG8 and REG9 respectively, and save the count value of the counter; S10, repeat S7 to S9 until the input of the pulse compression signal is completed; at this time, the values of REG7, REG8 and REG9 are the peak value of the pulse compression signal and the values of the two points adjacent to the peak value, at the same time, the count value output by the counter is regarded as the integer part of the channel fixed time delay.

8. The channel fixed time delay error estimation method of a wideband digital array radar according to claim 7, characterized in that, In step three, a polynomial is fitted using the peak point and the two points before and after the peak point, and the abscissa of the peak point of the polynomial is recorded as the fractional part of the channel fixed time delay, comprising: S11, record the values REG7, REG8 and REG9 output by the peak value estimation module as (0, f(0)), (-1, f(-1)) and (1, f(1)) respectively; S12, according to the above three points fitting such as f(t) = at 2 +bt+c, and calculate the abscissa of the peak value as follows: where t decimal is the fractional part of the channel fixed delay estimate.

9. The channel fixed time delay error estimation method of a wideband digital array radar according to claim 8, characterized in that, In step four, the fixed time delay error between each channel fixed time delay and the reference channel is obtained according to the following formula: wherein, is the estimated fixed delay error between the respective channel fixed delay and the reference channel.

Images