A signal processing method for orthogonal waveform design of a multi-station radar based on space-time code-phase code

Abstract

The application discloses a signal processing method of a multi-station radar orthogonal waveform design based on space-time code-phase code, and belongs to the field of signal processing. The signal processing procedure comprises space-time code demodulation, matched filtering, pulse accumulation, MTI and MTD. Each radar performs space-time code demodulation on the received echo signal, so that each pulse coefficient of the self radar echo is 1, and the sum of the pulse coefficients of the other radar echo for N (a positive integer power of 2) continuous pulses is 0. Then, subsequent matched filtering is performed, and each N pulse is accumulated, so that the echo from the other radar is cancelled, the interference from the other radar echo is greatly reduced, the processing mode does not affect the speed measurement accuracy, but affects the speed measurement range. The influence on the speed measurement range can be ignored by designing a space-time code matrix and a pulse accumulation mode, so that good anti-interference capability is achieved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of distributed multi-station radar signal processing, and particularly relates to waveform design and signal processing technology. BACKGROUND

[0002] In a distributed multi-station radar system, the ranging and velocity measurement of a moving target is completed, and the signal interference between adjacent radar stations is an important factor restricting the target detection performance of the multi-station radar. Therefore, the suppression of mutual interference between stations is an important problem that needs to be solved urgently in the multi-station radar system. The orthogonal waveform is commonly used to suppress mutual interference between stations. Time division orthogonal and frequency division orthogonal are limited in the utilization rate of time and frequency resources, and code division orthogonal has better resource utilization rate. Code division orthogonal waveform requires the signal to have good orthogonal characteristics, that is, the smaller the cross-correlation peak value, the better. For the linear frequency modulation signal and the phase coding signal commonly used by radars, the linear frequency modulation signal can obtain good autocorrelation characteristics by windowing, but its cross-correlation is poor, so it is not suitable for multi-station radar systems. At the same time, the phase coding signals with good anti-interference ability reported in the public literature, such as m sequence, gold sequence, etc., have unsatisfactory cross-correlation characteristics. Therefore, various coding optimization algorithms for code division orthogonal signals are emerging, but there is still room for optimization and improvement. Therefore, it is necessary to design a new waveform and corresponding signal processing method to solve the interference between adjacent radars.

[0003] In a multi-station radar system, each radar transmits different signals at the same time, and any radar will receive target echoes and direct wave interference from its adjacent radars. Therefore, the received signal can be understood as the radar's own echo signal and the interference signal of the adjacent radar station. After the radar's own echo signal is processed by matched filtering, a clear main lobe appears at the target position, and the interference signal of the adjacent radar station appears as a side lobe after matched filtering, thereby raising the side lobe of the received signal after matched filtering. Therefore, the peak side lobe ratio is selected as the measurement standard of target detection performance. The cross-correlation peak value of a signal with poor cross-correlation characteristics is higher, which makes the side lobe of the received signal after matched filtering higher, and the peak side lobe ratio lower, resulting in problems such as target being submerged or false target. Therefore, under the premise of not affecting the velocity measurement performance, improving the cross-correlation characteristics between signals and reducing the interference from adjacent radars to obtain a higher peak side lobe ratio can achieve accurate detection of the target.

[0004] The linear frequency modulation signal, the nonlinear frequency modulation signal modulation function is continuous, belongs to the "continuous type" signal, and the phase encoding signal is discrete and has a finite state, and belongs to the "discrete type" coded pulse compression signal. Since the phase encoding adopts a pseudo-random sequence, such a signal is also called a pseudo-random coded signal. The phase encoding signal is to decompose a wide pulse with a pulse width T into N sub-pulses with a width of tau. The common form of phase encoding is binary phase encoding, and the phase of each sub-pulse can be selected as 0 or pi radian, which is easy to obtain, but the correlation between different binary codes is not very ideal, and there is a lot of optimization space.

[0005] The space-time coding (STC) technology is used to suppress the influence of inter-correlation between waveforms. The core is to construct a space-time code matrix and a demodulation mode, to use space-time coding to decouple a group of complex waveform joint optimization problems into a series of independent waveform design problems, so that the existing single-channel waveform design results can be directly used, such as m sequence, gold code, Barker code and optimized gold code, and the coded signals with good correlation performance.

[0006] In a multi-station radar system, after the received target echo signal of each radar is matched filtered, the peak-to-sidelobe ratio of the self-correlation function of the transmitted signal is close to the peak-to-sidelobe ratio in an ideal state, but due to the existence of adjacent radar station interference, the detection performance of the target is affected by the inter-correlation characteristics between signals, and after the interference signal is matched filtered, it is superimposed on the self-correlation function of the target signal in the form of a sidelobe, resulting in a decrease in the peak-to-sidelobe ratio, and the target is not easy to detect. In order to solve this problem, a space-time code-phase code based orthogonal waveform design and signal processing method is proposed. SUMMARY

[0007] The technical problem to be solved by the present application is that common signals for distributed multi-station radar systems, such as m sequence, gold sequence, optimized phase code and two-dimensional code, are greatly affected by inter-correlation when detecting moving targets. Although the optimized binary code has greatly improved the autocorrelation and cross-correlation characteristics compared with the original binary code, in a multi-station radar system, after receiving the interference superposition of multiple adjacent radars, the peak-to-sidelobe ratio after matched filtering is greatly deteriorated, and cannot reach the peak-to-sidelobe ratio of the autocorrelation function of the optimized binary code.

[0008] In order to still maintain the autocorrelation peak-to-sidelobe ratio close to that of the optimized binary code after matched filtering in the presence of adjacent radar station interference, the present application proposes a space-time code-phase code based multi-station radar orthogonal waveform design signal processing method, which comprises the following steps:

[0009] Step 1: According to the number of radars N in the multi-station radar system and the number of pulses M transmitted by the radar in one frame, construct an N*M space-time code matrix A=[a1,a2,…,a i ,…,a N ] T , where a i =[g i,1 ,g i,2 ,…,g i,j ,…,g i,M ] T , 1≤i≤N, represents the space-time code sequence corresponding to the i-th radar, g i,j represents the space-time code corresponding to the jth pulse emitted by the i-th radar; secondly, the pulse accumulation number L needs to be determined. L must be the kth power of 2, and k is any positive integer, so as to construct the N*L matrix AL=[a1,a2,...,a i ,…,a N ] T , where a i =[g i,1 ,g i,2 ,…,g i,L ] T , the requirements that the matrix AL needs to meet are that the value of each code element in AL is 1 or -1, and then take a i , 1≤i≤N multiplied by each row element in AL to obtain the matrix ALD=[a1.*a i ,a2.*a i ,...,a N .*a i ] T , so that when L columns in the ALD matrix are added, only one row of the result is not 0, and the sum of the other rows is 0. N When all the above operations meet the requirements, the matrix AL is constructed; the matrix AL is then copied and connected to construct an N*M space-time code matrix, where M>=L;

[0010] Step 2: Modulate the radar transmit signal according to the constructed space-time code matrix A;

[0011] The matrix of binary phase code signal transmitted in multi-station radar system is S=[s1,…,s n ,…,s N ] T , where s n =[c n,1 ,…,c n,M ], the i-th radar, the j-th pulse is multiplied by A(i,j), where i=1,2,...,N, j=1,2,...,M, the modulated signal matrix Each station radar sends a modulated signal;

[0012] Step 3: Radar station i receives N radar station M echo pulses, and obtains M*K echo pulse matrix Re, where K is the number of sampling points in the pulse repetition period, demodulates the received signal, and multiplies the nth row in the Re matrix by A(i, n), where n = 1, 2, …, M, to obtain the demodulated echo matrix ReD;

[0013] Step 4: Signal processing;

[0014] The received continuous L pulses are accumulated, the interference of adjacent radar stations is eliminated after accumulation, only the autocorrelation of the self signal is retained, and the last obtained pulse signal is subjected to motion target detection to obtain the information of the motion target.

[0015] Further, the method for accumulating the continuous L pulses in step 4 is one of the following three ways:

[0016] The first accumulation method is to accumulate every L continuous pulses, and the starting interval of each accumulation pulse is L, and finally M / L pulses are obtained;

[0017] The second accumulation method is to accumulate every L continuous pulses, and the starting interval of each accumulation pulse is L / 2, and finally 2*M / L-1 pulses are obtained;

[0018] The third accumulation method is to accumulate every L continuous pulses, and the starting interval of each accumulation pulse is 1, and finally M-L+1 pulses are obtained.

[0019] Further, the principle analysis of the interference suppression of the signal processing method only considers that the system in step 4 contains three radars respectively sending binary code signals s1, s2, s3, and only transmits two pulses, and the space-time code matrix is designed as Therefore, the first pulse and the second pulse signals received by radar 1 are r1(t) = g1s1(t-τ1)-g2s2(t-τ2)+g3s3(t-τ3), r2(t) = g1s1(t-τ1)+g2s2(t-τ2)-g3s3(t-τ3), where g1, g2, g3 represent the path attenuation of the pulses transmitted by the three radar stations respectively, s1(.), s2(.), s3(.) represent three different binary code signals transmitted by the three radar stations, τ1, τ2, τ3 represent the time delay of the echo signals received by radar station 1 from the three radars, the demodulated echo matrix is matched filtered, and the matched filtering results of the first pulse and the second pulse signals are y1 = g1R1(τ1)-g2R 1,2 (τ2)+g3R 1,3 (τ3), y2 = g1R1(τ1)+g2R 1,2 (τ2)-g3R 1,3 (τ3), R1(τ1) represents the autocorrelation of s1, R 1,2(τ2) represents the cross-correlation of s1 and s2, R 1,3 (τ3) represents the cross-correlation of s1 and s3; adding the two can eliminate the interference of adjacent radar stations, and only the autocorrelation of the signal of radar station 1 is reserved, y1+y2=2g1R1(τ1);

[0020] Further, the motion target detection method in step 4 is an MTI or MTD method.

[0021] The beneficial effects of the present application are:

[0022] Without affecting the signal speed measurement accuracy, the interference of adjacent radar stations can be well suppressed, and the optimized two-phase code can fully exert the better autocorrelation and cross-correlation characteristics. The peak side lobe ratio of an optimized 512-bit M sequence is 27.3db. When not using space-time code for processing, the peak side lobe ratio of the received signal after matching filtering is reduced to 16db. When using space-time code for processing, the peak side lobe ratio of the received signal after matching filtering is 25.88db, close to the peak side lobe ratio of the autocorrelation function.

[0023] The common pulse radar mode is an accumulation mode, which reduces the speed measurement range to 1 / L. The space-time code design of the present application not only makes decoding simpler, but also makes the pulse accumulation mode more flexible. Pulse accumulation mode two can reduce the reduction of the speed measurement range, and pulse accumulation mode three makes the speed measurement range almost the same as the original.

[0024] At the same time, the more radar stations in the multi-station radar system, the more space-time code groups are needed. The number of pulse accumulations L can be increased, so that more space-time codes meeting the conditions are designed. The problem of speed measurement range reduction caused by increasing the number of pulse accumulations can also be suppressed by different pulse accumulation modes. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 It is a block diagram of orthogonal waveform design and signal processing technology based on space-time code-phase code.

[0026] Figure 2 It is a space-time-two-phase code modulation mode diagram.

[0027] Figure 3 It is a schematic diagram of the autocorrelation function of the optimized M sequence.

[0028] Figure 4 It is a schematic diagram of the cross-correlation function between the second M sequence and the first M sequence in the three optimized M sequences.

[0029] Figure 5Figures of comparison between echo signal matched filtering results of not using space-time processing and using space-time processing; (a) is only two-phase code matched filtering result, (b) is space-time-two-phase code matched filtering result.

[0030] Figure 6 Figure of speed measurement precision and range of only two-phase code signal.

[0031] Figure 7 Figure of three pulse accumulation modes.

[0032] Figure 8 Figure of influence of different pulse accumulation modes on speed measurement of moving target; (a) pulse accumulation mode one, (b) pulse accumulation mode two, (c) pulse accumulation mode three. DETAILED DESCRIPTION

[0033] Step 1: let the center frequency of the transmitted space-time-two-phase code signal be 4.9 GHz, the bandwidth be 100 MHz, the pulse repetition interval be 2.5 ms, and there be 80 pulses in one frame. Let the multi-station radar system include 3 radar stations, and 3 optimized 512-bit M sequences be used as the transmitted signals of the respective radars, thereby constructing a 3*80 transmitted signal matrix S. Let the pulse accumulation number be 4, and a 3*4 space-time matrix be first constructed Copy the AL to connect 20 times to construct a 3*80 space-time code matrix A, and obtain a transmitted signal matrix Sig=A.*S. The construction principle is shown in Figure 2 The adjacent radar station spacing is 1000 m, a single target scene is simulated, and clutter is not considered. The target distance from radar station 2 is 1000 m, and the radial velocity is 1 m / s. Radar station 2 receives the signals of the 3 radar stations, and the simulation verification is performed from the perspective of radar station 2.

[0034] Step 2: compare the peak-to-sidelobe ratio results of only using two-phase code signal P1 and using space-time-two-phase code signal P2 after pulse compression. The peak-to-sidelobe ratio of the optimized second M sequence (i.e. the phase code transmitted by radar station 2) is shown in Figure 3 as 27.3 db. The cross-correlation of the first M sequence (i.e. the signal transmitted by the first radar station) and the second M sequence is shown in Figure 4 , which can reach -21.3 db. Radar station 2 sends signal P1, receives the signals of the 3 radar stations, and then performs matched filtering. The peak-to-sidelobe ratio is shown in Figure 5 (a), which is deteriorated to 16 db. Radar station 2 sends signal P2, receives the signals of the 3 radar stations, and then performs matched filtering. The peak-to-sidelobe ratio is shown in Figure 5(b) as shown, still can reach 25.8db close to the second M sequence autocorrelation peak side lobe ratio. The two-phase code optimization algorithm mentioned in the public literature, such as CAN algorithm, iterative block coordinate descent algorithm, etc., by optimizing the two-phase code, the autocorrelation peak side lobe ratio of the 512-bit M sequence is about 27db, and the cross-correlation peak value can be reduced to about -21db, but considering the interference of receiving multiple radar signals, the peak side lobe ratio after matching filtering will be reduced to about 17db, which cannot reach the autocorrelation peak side lobe ratio. From Figure 3 、 Figure 4 and Figure 5 (a) (b) from the results, the method can almost eliminate the interference of adjacent radar stations, so that the peak side lobe ratio after matching filtering is close to the peak side lobe ratio of the optimized M sequence, and the interference suppression is greatly improved.

[0035] Step 3: The echo signal received by the simulated radar station 2 (only using two-phase code signal) is added with time delay and Doppler frequency according to the distance and radial velocity of the moving target in the simulated single target environment. The interference of adjacent radar stations is considered as direct wave interference, so the Doppler frequency of the adjacent radar station interference signal is 0. Radar station 2 receives the signals of 3 radar stations, and after matching filtering, MTI and MTD processing, the Doppler profile of the distance unit where the target is located is as shown in Figure 6 The distance resolution is 1.5m, the velocity resolution is 0.15m / s, and the velocity measurement range is -6~6m / s. The detection result is consistent with the calculation result. Figure 6

[0036] Step 4: The echo signal received by the simulated radar station 2 (using space-time two-phase code signal) is added with time delay and Doppler frequency according to the distance and radial velocity of the moving target in the simulated single target environment. The interference of adjacent radar stations is considered as direct wave interference, so the Doppler frequency of the adjacent radar station interference signal is 0. Radar station 2 receives the signals of 3 radar stations, and after space-time code demodulation, matching filtering, pulse accumulation, MTI and MTD processing.

[0037] Step 5: The signal processing procedure mentioned in step 4 is used to demodulate the space-time code of the signal received by radar station 2. The space-time code corresponding to radar station 2 is the second row a2 = [1 -1 1-11-1…] in matrix A. During demodulation, only the received 80 pulses are multiplied by the corresponding code elements in a2 in order, which makes the coefficients of the pulses transmitted by itself all be 1, which will not be eliminated during accumulation, and the coefficients of the pulses transmitted by other radars are all 0 every 4 continuous pulses. This process is equivalent to modifying the original space-time code matrix A to After demodulation, matching filtering is performed. ​

[0038] Step 6: After demodulating and matching the signal received by radar station 2 in step 5, the next step is to perform pulse accumulation to eliminate interference. There are three pulse accumulation methods: Figure 7 As shown in the figure, three different accumulation methods are used for processing. It is known that a frame signal emitted by the radar contains 80 pulses, and the pulse radar accumulation number is 4. After processing with accumulation method 1, 20 pulses are obtained, after processing with accumulation method 2, 39 pulses are obtained, and after processing with accumulation method 3, 77 pulses are obtained. Subsequent MTI and MTD processing are then performed to intercept the Doppler profile of the range unit where the target is located. Figure 8 (a) is the Doppler profile after signal processing in accumulation mode 1. Figure 8 (b) is the Doppler profile after signal processing in accumulation mode 2. Figure 8 (c) is the Doppler profile after signal processing in accumulation mode 3. Figure 8 The analysis of (a), (b), and (c) shows three processing methods. The target is located in the same Doppler channel, which means that the speed measurement accuracy is not affected by the pulse accumulation method. However, the speed measurement range is reduced compared to the method that only uses binary code without pulse accumulation. However, this problem can be well solved by changing the accumulation method, such as accumulation method 3.

[0039] After experimental verification, this method can effectively suppress the interference of adjacent radar station signals, so that the radar can give full play to the excellent correlation characteristics of the optimized binary code. At the same time, this method will not affect the speed measurement accuracy of the radar. The problem of reduced speed measurement range can also be suppressed as much as possible by changing the pulse accumulation method. Therefore, this method solves the problem of interference from adjacent radar stations affecting target detection in a distributed multi-station radar system while ensuring that the radar detection accuracy meets the requirements.

Claims

1. A signal processing method for multi-station radar orthogonal waveform design based on space-time code-phase code, the method comprising: Step 1: According to the number of radars N in the multi-station radar system and the number of pulses M transmitted by the radar in one frame, construct an N*M space-time code matrix A=[a1,a2,…,a i ,…,a N ] T , where a i =[g i,1 ,g i,2 ,…,g i,j ,…,g i,M ] T , 1≤i≤N, represents the space-time code sequence corresponding to the i-th radar, g i,j represents the space-time code corresponding to the jth pulse emitted by the i-th radar; secondly, the pulse accumulation number L needs to be determined. L must be the kth power of 2, and k is any positive integer, so as to construct the N*L matrix AL=[a1,a2,...,a i ,…,a N ] T , where a i =[g i,1 ,g i,2 ,…,g i,L ] T , the requirements that the matrix AL needs to meet are that the value of each code element in AL is 1 or -1, and then take a i Multiply each row element in AL to get the matrix ALD = [a1.*a i ,a2.*a i ,...,a N .*a i ] T , so that when L columns in the ALD matrix are added, only one row of the result is not 0, and the sum of the other rows is 0. N When all the above operations meet the requirements, the matrix AL is constructed; the matrix AL is then copied and connected to construct an N*M space-time code matrix A, where M>=L; Step 2: Modulate the radar transmit signal according to the constructed space-time code matrix A; The matrix of binary phase code signal transmitted in multi-radar system is S=[s1,…,s i ,…,s N ] T , where s i =[c i,1 ,…,c i,M ] represents the M pulses emitted by the i-th radar, and the binary code signal matrix S is multiplied by the space-time code matrix A to obtain the modulated signal matrix Each radar sends a modulated signal; Step 3: Radar i receives M echo pulses from N radars and obtains an M*K echo pulse matrix Re, where K is the number of sampling points in the pulse repetition period. The received signal is demodulated and the nth row in the Re matrix is ​​multiplied by A(i,n), where n = 1, 2, …, M, to obtain the demodulated echo matrix ReD. Step 4: Signal processing; The received L consecutive pulses are accumulated, and the interference of adjacent radars is eliminated after accumulation, and only the autocorrelation of the own signal is retained. The final pulse signal is used for moving target detection to obtain the information of the moving target.

2. The signal processing method for multi-station radar orthogonal waveform design based on space-time code-phase code according to claim 1, characterized in that: The method for accumulating L consecutive pulses in step 4 is one of the following three methods: Accumulation mode 1: every L consecutive pulses are accumulated, the starting interval of each accumulated pulse is L, and finally M / L pulses are obtained; Accumulation mode 2: every L consecutive pulses are accumulated, the starting interval of each accumulated pulse is L / 2, and finally 2*M / L-1 pulses are obtained; Accumulation mode three: every L consecutive pulses are accumulated, the starting interval of each accumulated pulse is 1, and finally M-L+1 pulses are obtained.

3. The signal processing method for multi-station radar orthogonal waveform design based on space-time code-phase code according to claim 1, characterized in that: From the principle analysis of signal processing to suppress interference, we only consider the system in step 4, which contains three radars that send two-phase code signals s1, s2, and s3 respectively, and only transmit two pulses. Then the space-time code matrix is ​​designed as Therefore, the first and second pulse signals received by radar 1 are r1(t)=g1s1(t-τ1)-g2s2(t-τ2)+g3s3(t-τ3); r2(t)=g1s1(t-τ1)+g2s2(t-τ2)-g3s3(t-τ3), where g1, g2, and g3 represent the path attenuation of the pulses transmitted by the three radars, s1(.), s2(.), and s3(.) represent the three different binary phase code signals transmitted by the three radars, and τ1, τ2, and τ3 represent the echo delays received by radar 1 from the signals transmitted by the three radars. The demodulated echo matrix is ​​matched filtered, and the matched filtering results for the first and second pulse signals are: y1=g1R1(τ1)-g2R 1,2 (τ2)+g3R 1,3 (t3); y2=g1R1(τ1)+g2R 1,2 (τ2)-g3R 1,3 (t3); R1(τ1) represents the autocorrelation of s1, R 1,2 (τ2) represents the cross-correlation between s1 and s2, R 1,3 (τ3) represents the cross-correlation between s1 and s3; adding the two together eliminates the interference from the adjacent radar, leaving only the autocorrelation of radar 1's own signal, y1+y2=2g1R1(τ1).

4. The signal processing method for multi-station radar orthogonal waveform design based on space-time code-phase code according to claim 1, characterized in that: The moving target detection method in step 4 is the MTI or MTD method.

Images