JPH0330112B2 - - Google Patents

Description

[Detailed description of the invention]

(Technical Field) The present invention relates to a pulse compression radar that uses transmission pulses phase-modulated with an uncorrelated binary arbitrary random number code sequence of 1 and -1. (Prior Art) Fundamentally important performances for radar are high target detection performance and high target location performance. However, these two elements become contradictory requirements when designing a transmission waveform. That is, the former requires a long transmission pulse width, and the latter requires a short transmission pulse width. Pulse compression technology was devised to simultaneously satisfy these contradictory demands. According to this technology, the reflected waves from the target of radio waves transmitted with a long transmission pulse width are processed after reception and compressed into a short pulse width without reducing the average power, resulting in high detection performance and a long range. It also satisfies the orientation performance. This pulse compression technology is based on the modulation method of the transmitted pulse. (1) Linear frequency modulation method (for example, a method in which the frequency of the carrier wave in the transmitted pulse is increased linearly from the front end to the rear end of the transmitted pulse) (2) Non-linear frequency modulation method (for example, a method in which the frequency of the carrier wave within a transmission pulse increases linearly from the front end to the rear end of the transmission pulse) (3) Coded phase modulation method (for example, (methods in which the phase of the carrier wave within a transmitted pulse is changed in accordance with a specific code sequence), and each method has been put into practical use by taking advantage of its characteristics. Currently, linear frequency modulation is widely used due to its ease of generating modulated signals and manufacturing devices for pulse compression, but coded phase modulation, which is resistant to mutual interference and interference, will also be widely used in the future. There is a tendency to The coded phase modulation method divides the transmitted pulse into several small parts (N time cells) and performs phase modulation corresponding to the code for each time cell.Currently, it mainly uses 0 Binary phase modulation of 180° and 180° is used. Furthermore, various code sequences have been proposed for the arrangement order of the phases of each time cell, but among these, the main one from an applied perspective is a sequence composed of binary codes of 1 and -1. (1) Barker series (2) M series, etc. However, in the Barker series, the maximum value of the number of time cells N is 13 (at that time, the peak-to-sidelobe ratio = -
22.3dB), it is not possible to achieve a high pulse compression ratio. FIG. 3A shows the state of pulse compression in the case of the Barker series. Furthermore, the M sequence uses a code sequence with N numbers as one cycle, but the length N of the random number sequence cannot be arbitrarily selected (N=2 ⁿ -1, n=1, 2, 3,...) Furthermore, even if the length of the sequence is determined, the type of sequence is limited by the value of N (2 types for N = 15, 6 types for N = 31, 6 types for N = 63, etc.). The lobe ratio is approximately −10 logN dB), and even when used for pulse compression, conventional methods cannot sufficiently suppress time side lobes. FIG. 3B shows the state of pulse compression in the case of M sequence. (Objective of the Invention) The present invention eliminates the drawbacks of the conventional encoded phase modulation pulse compression method using a code sequence, and further expands the advantages. The objective is to provide a pulse compression radar that causes less mutual interference during operation and is less susceptible to intentional interference. (Structure of the Invention) In the present invention, in the modulation within the transmission pulse, 1 and -1, which are completely uncorrelated, are not used for a specific code sequence.
A binary random number sequence of arbitrary length consisting of is used, and the ratio of the signal peak output level to the average value of the time sidelobe level is maximized by using a pulse compression filter that gives the received signal optimal weighting determined by the transmission code sequence. Pulse compression is performed so that the received signal of multiple pulses after pulse compression is subjected to FFT
By performing coherent integration processing such as (Fast Fourier Transform) processing, the power of time side lobes that randomly occur for each pulse is dispersed in the frequency domain, and the time side lobe components output for each filter bank are reduced. is being carried out. That is, phase modulation within the transmitted pulse is performed using an uncorrelated binary random number sequence of an arbitrarily selected length,
An example of pulse compression without optimal weighting is shown in FIG. 4A, and the time sidelobe level is higher than that of an example of pulse compression using the M sequence shown in FIG. 3B. On the other hand, when pulse compression is performed using a pulse compression filter weighted by calculating the optimal weighting coefficient, the improvement is improved as shown in Figure 4B, and the pulse compression output is further improved by utilizing the uncorrelation of the code sequence for each transmitted pulse. Figure 4C shows the result of coherent integration by Fourier transform over multiple pulses for each range bin, and the sidelobe level has been sufficiently reduced. In addition, in the encoded phase modulation method of transmitted pulses adopted in the present invention, the modulated waves are uncorrelated for each transmitted pulse, so even if many radars are operated simultaneously in the same frequency band, the pulse compression of each radar will be reduced. A filter can match and correlate only the reflected signal of its own transmitted radio waves, but it cannot correlate with incoming signals from other radars, so the output power level of the filter is lower than the reflected signal of its own transmitted waves. 1/N, and mutual interference is significantly reduced. Furthermore, for the same reason, it is highly resistant to intentional interference signals. This method of transmitting and receiving uncorrelated code sequences for each transmission pulse has superior features compared to Barker sequences and M sequences. (Embodiments of the Invention) Hereinafter, embodiments of the pulse compression radar according to the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of an embodiment of the present invention. As shown in this figure, the pulse compression radar consists of a transmitting device 1, an antenna 2, a receiving device 3, a signal processing device 4, and a display/following device 5. In the transmitter 1, the random number generator 11 has 1, -
It consists of a circuit that generates N binary random numbers of 1 at time intervals of the time cell width, or a recording device that stores random numbers created in advance and a reading circuit thereof. The COHO oscillator 13 is a stable intermediate frequency oscillator that serves as the standard for this radar, and its output is
The COHO signal is sent to the random number generator 11 by the phase modulator 14.
It undergoes 0° or 180° phase modulation by a binary random number sequence signal. This output is converted to local oscillator 1 by frequency converter 15.
The transmitting/receiving switch 1 is mixed with the carrier wave generated in 6, converted to a high frequency, and amplified by a power amplifier 17.
8 and is transmitted from antenna 2. On the other hand, the reflected signal from the target received by the antenna 2 is sent to the frequency converter 31 of the receiving device 3 via the transmitting/receiving switch 18, where it is converted to an intermediate frequency, amplified by the intermediate frequency amplifier 32, and then passed through the phase detector. 33 and 34. The phase detectors 33 and 34 are
The phases of the COHO signal and the signal obtained by delaying the COHO signal by a 90° phase shifter 35 and the received intermediate frequency signal are compared and converted into an I channel video signal and a Q channel video signal, respectively. The video signals of the I and Q channels enter pulse compression units 41 and 42, respectively, which constitute a pulse compression filter, where the correlation with the transmission code sequence is taken and pulse compression is performed. As shown in FIG. 2, the pulse compressors 41 and 42 include an N-bit standby register 61 corresponding to N range bins, and a tapped delay line 62 consisting of N time delay elements corresponding to N range bins. , N multipliers 63 and adders 64. The weight calculation unit 12 in FIG. 1 calculates the optimum weighting coefficient determined from the code sequence for each transmission pulse, or stores the calculated result in advance, and sends this to the pulse compression units 41 and 42. This weighting factor is the second
It is stored in the standby register 61 shown in the figure. On the other hand, the I or Q channel video signal sent from the receiving device 3 of FIG. 1 is sequentially input to the tapped delay line 62. As mentioned above, both the standby register 61 and the tapped delay line 62 are composed of N cells, and each cell is connected to an individual multiplier 63. Now, the received pulses obtained from one reflecting object are sequentially input to the tapped delay line 62 as a video signal expressed by equation (1). Ys(n)=EsV(nΔt)≡EsV(n)...(1) However, ES: Peak voltage of the signal, V(n): +1 or -1 occurs for each time cell width Δt with a probability of 0.5. The code sequence is n=1, 2,...,N. At this time, when the optimal filter coefficient Wo(i) obtained by multiplying the time replica of the transmission code sequence by the weighting coefficient W(i) is given to the standby register 61, the multiplier 63 of each tap gives The output will be obtained. T(i)=Es・V(l−i+1)・Wo(i)……(2) However, Wo(i)10V(N−i+1)・W(i)……(3) l=1, 2,..., 2N-1 are each adder 6
4 represents the timing number at which the compressed video signal is output. Specifically, l = 1 is when the first time cell of the received pulse is input to tap i = 1 of the tapped delay line 62, and When the time cell of is input to tap i=N, l=2N-1
It is. As a result, the peak power of the signal from the adder 64 (when l=N) is: S=| _N 〓 ⁱ⁼¹ T(i)| ² =E _S ² | _N 〓 ⁱ =1 )} ² W(i)｜ ² =E _S ² ｜ _N 〓 ⁱ⁼¹ W(i)｜ ² E _S ² {[W(1)W(2)...W(N)]1 1 〓 1} ² = E _S ² ｜〓 ^t 〓｜ ² ...(4) However, 〓 ^t = [W(1)W(2)...W(N)]: N-element vector representing the optimal weighting coefficient, 〓 ^t = [1 1...1]: N-element vector with all elements being 1, t: Indicates transposition (for example, vector [a ₁ a ₂ ...a _N ]
The transposed vector of is a ₁ a ₂ 〓 a _o . The rows and columns have been swapped. Moreover, the average power of the time sidelobe is given by equation (5), if the power of minute DC components is ignored. That is, the signal output from the adder 64 is l
At the timing of =1 to 2N-1 (l≠N), a time sidelobe is output in the form of noise, and at l=N, a peak signal whose time width is compressed from NΔT to ΔT is output. This can be expressed as a vector as follows.

【table】 ~
~
X^
W ₀

[Table] =X^V^W=X^W ₀
Therefore, the average time sidelobe power is l=1
Because of the additive average of time sidelobe power at ~2N−1 (l≠N), P _av = 1/2N−2E _S ² { _N−1 〓 ^l=1 | V ₁ (l) | ² + _2N−1 〓 ^l=N+1 ｜V ₂ (l)｜ ² }=E _S ² /2(N-1) { _N-1 〓 ^l=1 ｜〓(l)〓 ₀ ｜ ² + _2N-1 〓 ^{l= N+1} ｜〓(l)〓 ₀ ｜ ² }=E _S ² /2 (N-1) { _N-1 〓 ^l=1 〓 ₀ ^t 〓 ^t (l)〓(l)〓 ₀ + _2N-1 〓 ^l=N+1 ₀ ^t 〓(l)〓(l)〓 ₀ }=E _S ² /2(N-1) {〓 ₀ ^t [ _N-1 〓 ^l=1 〓 ₁ (l)]〓 ₀ +〓 ₀ ^t [ _N-1 〓 ^l=1 ₂ (l)]〓 ₀ }E _S ² /2 (N-1) {〓 ^t [ _2N-1 〓 ^l=N+1 〓 ₁ (l)]〓 +〓 ^t [ _2N-1 〓 ^l=N+1 ₂ (l)]〓}=E _S ² 〓 ^t [1/2 (N-1) { _N-1 〓 ^l=1 〓 ₁ (l)+ _{2N -1} 〓 ^l=N+1 ₂ (l)〓〓]=E _S ² 〓 ^t 〓〓 ……(5) However, the vector 〓(l) is the l-th row vector of the matrix 〓 shown on the previous page, and the matrix 〓 ₁ (l)
and 〓 ₂ (l) are both covariance matrices of vectors 〓 ^t (l) and 〓(l), and furthermore, in the above equation (5), |〓(l)〓 ₀ | ² = (〓〓) ^t [ 〓 ^t (l)〓(l)〓（〓〓）
=〓 ^t 〓 ^t [(〓 ^t (l)〓(l)] 〓〓＝〓 ^t [〓 ^t (l)〓(l)]〓 ^t 〓〓＝〓 ^t [〓 ^t 〓(
l)]〓. (∵ 〓 and [〓 ^t (l)〓(l)] are both symmetric matrices, so the order can be changed, and 〓 ^t
〓 becomes the unit matrix. ) 〓=1/2(N-1) { _N 〓 ⁱ⁼¹ 〓 ₁ (l)+ _2N-1 〓 ^l=N+1 〓 ₂ (l)} ...(6) 〓=[W(1) W(2)...W(N)] ^T ...(7) 〓=[1 1...1] ^T ...(8) 〓 ₂ (l)=[00...0 V(N) V(N-1)...V(l-
N+1)] ^T ...(10) (N+1l2N-1) T: Transpose 〓, 〓, 〓, 〓 ₁ (l), 〓 ₂ (l): N-element vector. Using the above analysis results, the signal peak power S
and the average power P _av of the time sidelobe α=S/P _av =｜〓 ^t 〓｜ ² / ^t 〓〓 = ^t 〓〓／〓 ^t
Find the weighting coefficient vector = [W(1)W(2)...W(N)] ^t that maximizes 〓...(11) on average. however,

[Formula] is an N-element square matrix. Now, if we transform the above equation (11), ^t 〓〓=α〓 ^t 〓〓 Taking the variation on both sides, we get δ〓 ^t 〓〓+〓 ^t 〓δ〓=δα(〓 ^t 〓〓) +α(δ〓 ^t 〓〓+〓 ^t 〓δ〓) Here, δγ ² does not change with respect to the reduction change in δ〓, so δγ ² =0. Therefore, 2δ〓 ^t (〓〓−α〓〓)=0 must hold, and the following equation is finally derived. 〓〓−α〓〓=〓(〓 ^t 〓)−α〓〓=〓β−α〓
〓=0 (However, α and β are scalar quantities) Therefore, as the optimal weighting coefficient vector, 〓=β/α〓 ^-1 〓=μ〓 ^-1 〓=[W(1)W(2)...W( N)
] ^t ...(12) (where μ=β/α: any real scalar quantity that is not zero) is obtained. 〓 ^-1 : It is the inverse matrix of 〓. Therefore, the optimal filter coefficient Wo(i) given to the standby register 61 is a value determined by substituting equations (7) and (12) into equation (3). The optimal weighting coefficient determined above reduces the time sidelobe by minimizing the average timesidelobe level. As a result, the time sidelobe level of pulse compression in a certain binary code sequence used in Figure 4A is as shown in Figure 4B.
The figure shows how the time sidelobes are reduced so that the ratio of average power/peak signal power is minimized. The contents of each cell of the standby register 61 and the contents of each cell of the tapped delay line 62 storing the optimal filter coefficients obtained as described above are as follows.
N multipliers 6
The outputs of 3 are added in an adder 64 and sent as an I or Q channel compressed video signal to the FFT section (the section that performs Fast Fourier Transform) 4 in FIG.
Sent to 3. The FFT unit 43 collects several pulses for each range bin of the pulse compression signal as shown in FIG. 4B (over multiple sweeps of the radar).
Coherent integration is performed, and the outputs are synthesized in an IQ synthesis section 44, and then a target detection section 45 detects the output of the frequency bank in which the output amplitude has a maximum value for each range bin. At this time, if necessary, the output of the frequency bank whose Doppler frequency corresponds to 0 is cut off to eliminate ground reflections and the like. The video signal that has undergone all the above processing is sent to the display/tracking device 5 and used for display or tracking. (Effects of the Invention) As described above, according to the pulse compression radar of the present invention, due to the good suppression performance against time side lobes and the uncorrelation (secrecy) between transmitted pulses, (1) prediction of strong clutter is achieved; (2) Effective use of frequencies (radio wave resources) through mutual interference removal function when multiple radars are operating simultaneously; (3) Operation in situations where intentional interference is expected. ,
A significant effect can be expected.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of a pulse compression radar according to the present invention, FIG. 2 is a block diagram showing an example of a pulse compression section according to the present invention, and FIG. 3 is a block diagram showing an example of a pulse compression section according to the present invention. Graph showing an example of the result. FIG. 4 is a graph showing an example of the result of encoded pulse compression according to the present invention. DESCRIPTION OF SYMBOLS 1... Transmitting device, 2... Antenna, 3... Receiving device, 4... Signal processing device, 5... Display device/following device, 11... Random number generation section, 12... Weight calculation section, 13... COHO oscillator, 14... Phase modulator, 15... Frequency converter, 16... Local oscillator, 17... Power amplifier, 18... Transmission/reception switch,
31... Frequency converter, 32... Intermediate frequency amplifier, 33, 34... Phase detector, 35... 90° phase shifter, 41, 42... Pulse compression section, 43...
FFT section, 44... IQ synthesis section, 45... Target detection section, 61... Standby register, 62... Tapped delay line, 63... Multiplier, 64... Adder.

Claims (1)

[Claims] 1 Phase-modulates the transmission pulse with an uncorrelated binary arbitrary random number code sequence of 1 and -1 and transmits it, and passes the received signal through a pulse compression filter with optimal weighting determined by the transmission code sequence to calculate the peak output level and time. Pulse compression is performed so that the ratio of the side lobe to the average output level is maximized, and coherent integration is performed by passing the output after pulse compression through a group of narrow band filters over multiple pulses for each range bin. , a pulse compression radar characterized by suppressing time sidelobes.