FR2746505A1 - Eliminating secondary lobe effects remotely for radar Doppler compression pulses - Google Patents

Abstract

The process includes using a correlator (COR), a Doppler filter (FD) and an integrator (INT), and consists of effecting a double detection w.r.t. the signal provided by the output of the integrator of the signal processing chain. A first detection with a high threshold allows detection of only the main peaks of large amplitude whose secondary lobes cause pollution of neighbouring signals. One then determines, for each of the distant gates, the global level of the secondary lobes (ST) and one subtracts this from the received signal (A) in each of the distant gates. The second detection, using a manual threshold, is then effects on a signals (A-ST), where the effects of the secondary lobes, associated with peaks of large amplitude, have been eliminated.

Description

 The present invention relates to a method and device for remote side lobe elimination due to pulse compression.

 In radars intended to detect and locate mobile vehicles on the ground, it is necessary to operate short and spaced probe shots in order to reduce the possibility of destruction by anti-radiation missiles and to make it more difficult to opposing countermeasures. The echoes of vehicles must therefore be detected and located very quickly, which supposes the existence of automatic means to avoid the saturation of the operator by false alarms from clutter residues and active or passive interference signals. This indispensable automatic control of the false alarm rate must take into account the particular nature of the targets and the characteristics of the various parasitic signals.

 The false alarm is caused by the presence, in a resolution cell of the radar, of a parasitic signal exceeding the detection threshold. This parasitic signal can come either from a parasitic echo from an external source or from a secondary lobe in distance due to the compression of pulses and from a useful large amplitude signal located in a resolution cell close to the cell considered.

 The use of the pulse compression technique results in the appearance, after reception and correlation, of secondary lobes in distance, that is to say spurious signals having the same spectral distribution as the principal peak, or useful signal, but smaller amplitudes. The distance distribution as well as the amplitude of these sidelobes depend on the type of coding used, as well as reception circuits, correlator and weighting of secondary lobes. In the case, for example, of a code, called "Barker", at thirteen moments, associated with a weighting of the side lobes, any signal whose main peak after detection and correlation exceeds the normal noise level by thirty decibels. radar, leads to a significant probability of detecting side lobes.

 The present invention therefore proposes to provide a method for reducing this probability of detection, and thus making it possible to reduce the false alarm rate, and a device using this method.

où les contributions des pics principaux de grande amplitude sont limitées à celles des pics présents dans les P portes en distance entourant la porte en distance considérée, P avant et P après, (avec P nombre pair). According to the invention, this process consists for each of
N gates in distance of a given recurrence period, to evaluate the overall level of secondary lobes ST. in the distance index gate j and subtract it from the measured signal amplitude A.
J this gate in distance, to deduce a real amplitude of signal λ = Aj - STj not polluted by the sidelobes associated with major peaks Ak of large amplitude, this real amplitude A. being J compared to the manual threshold k0 of the radar so to determine if there is or there is no target in this gate at a distance. According to a development of the invention, this overall level of secondary lobes ST. in the door
J in index distance j is calculated from the contributions S of
jk major peaks of great amplitude Ak, detected in gates in distance of index k, according to the law of general composition

where the contributions of the main peaks of large amplitude are limited to those of the peaks present in the P gates at a distance surrounding the gate in the considered distance, P before and P after, (with P even number).

2 2
The invention will be better understood and other characteristics will become apparent with the aid of the description below and the attached drawings in which - FIG. 1 represents a radar signal processing chain.
having a device for eliminating secondary lobes
distance according to the present invention; FIG. 2 represents a temporal diagram allowing to understand
the principle of the elimination of side lobes according to the present
invention; and FIG. 3 represents a diagram of the device for eliminating lobes
secondary agents according to the present invention.

 FIG. 1 shows a conventional signal processing chain for Doppler radar with pulse compression. This chain comprises a COR correlator, a FD Doppler filter, an INT integrator detector and a S threshold system k0. This signal processing chain provides the threshold system S with signals which, in each of the remote gates, can be polluted by the effects of the main peaks of large amplitude, located in adjacent distance gates. According to the present invention, an EL device for eliminating these sidelobes is inserted in this processing chain, after the detector-integrator INT and in replacement or in complementary use with this threshold system S. This chain of processing of the radar signal receives, from the buffer memory MT, at time t. ., the components aij and bij of the sample Vij of the signal vector, received in the gate in distance of index j during the ileme period of recurrence.

The sampling and coding of the received signal vectors Vi (t) are obtained by the coder-samplers COD 1 and COD 2, respectively of the in-phase channel and of the quadrature channel. These encoders receive mixers MEL I and MEL 2 components a. (t) and b. (t) of this analog signal Vi (t) received during the second recurrence period.

Any signal whose principal peak, after correlation and integration, exceeds the normal noise level of the radar by thirty decibels, results in a significant probability of detecting its own side lobes on adjacent distance gates. Thus, in order to reduce this probability of detection and therefore to reduce the false alarm rate, the present invention proposes to perform, at the output of the integrator, a priori detection of
major peaks corresponding to useful targets whose level is
greater than a threshold k1 = ak0, where kg is the corresponding manual threshold
the normal sound of the radar; - to determine beforehand k1 from the following relation

<tb>#Prob.<SEP><SEP> of <SEP> detection <SEP> of <SEP> targets <SEP> for <SEP> signal <SEP>#<SEP> x <SEP> dB <SEP>#sur<SEP> k1 <SEP > =
<tb><SEP> noise
<tb><SEP>#Prob.<SEP><SEP> of <SEP> detection <SEP> of <SEP> targets <SEP> for <SEP> signal <SEP>#<SEP> 0 <SEP> dB <SEP> threshold <SEP>#Select<SEP> k0 <SEP>;<September>
<tb><SEP> noise
<tb> and, when a main amplitude peak i has been detected for the threshold k1, in the index distance gate k - to determine its contribution Sjk, at the detection threshold, in the PPs.
P doors surrounding this main peak, 2 front doors and 2 doors
after this number of P doors chosen depending on the rate of compres
radar. This contribution Sjk, in the door distance
of index j, is obtained from the expression
PP
Sjk = Ak - x dB with k - 2 # j # k + 2; - to evaluate, for the N gates in distance of the period of recurrence
considered, the value of the estimated overall level of secondary lobes ST.

J
in the gate in distance of index j. This threshold ST. is a function of
P each of the contributions Sjk of the peaks located within 2 doors of the door j considered. The law of general composition of these contri butions can be written

where N is the total number of doors and where I 4 n 4 2. The quadra law
tick (n = 2) is chosen if we assume that the main peaks
detected Ak, of great amplitude, are independent and therefore that the
relative phases of the sidelobes associated with any two of
these main peaks are random; - to subtract in each gate in distance the estimated level of lobes
secondary ST. in the gate in distance from index j to amplitude A.

J
of the signal in this door in distance. We obtain an amplitude A., J
after removal of the sidelobes, of the form
 = Aj - STj; - to compare the new value Â. from the main peak to the manual threshold ko
J
and to detect the echo on a value jj of main peak where the effects
secondary lobes due to the main peaks of adjacent gates
have been removed.

 The value of the attenuation x chosen is variable according to the chosen compression ratio. It can be equal to x = 24 dB for example in the case of a Barker code with thirteen moments with weighting.

c'est-à-dire STjn = (STj)n + (STj)n. However, if it is desired to determine for each door in distance the contribution of all the adjacent doors, located at
P less than 2 gates before or after, one obtains PN determinations, which requires an important material. The solution of the present invention to this problem can be understood by means of the time diagram shown in FIG. 2. This solution according to the present invention makes it possible to replace these PN determinations by 2N determinations plus N additions.
It makes it possible to obtain a reduction in the volume of the circuits and a simplification of these circuits. It corresponds to the decomposition of the expression STjn, giving the law of composition of the coefficients Sjkn, in two summations, one on the values of k lower than j and the other on the values of k greater than j, either

that is, STjn = (STj) n + (STj) n.

2 1
During the period of recurrence considered, the signals supplied in each remote gate, by the integrator-detector INT, are stored in a buffer memory so that they can be processed in two stages, subsequently.

 On the time diagram a) of FIG. 2, are represented the principal amplitude peaks Ak, detected for the threshold k1 at the output of the buffer of the detector-integrator and appearing during a recurrence period, cut out here. in N doors in distance.

For the first detected peak, which corresponds to the smallest distance, P has been represented his contributions S in the 2 doors located on the
jk P and other of this peak, or of the gate of index j = k - 2 at the
2 a gate of index j = k + 2. These contributions correspond to a chosen "rectangular" modeling of these distant side lobes all supposed to be located at x dB below the amplitude Ak of this detected main lobe.

 On the time diagram b) of Figure 2, are shown the first N determinations performed by reading the buffer of the integrator in the increasing direction of the remote gates.

As soon as there is detection, for the threshold k1 = &alpha; k0, of a main peak of amplitude Ak high, an intermediate memory will receive on the
P next distance gates, j = k + 1 to j = k + P, the attenuated value 2 2 Sjkn = (Ak - x) n, in decibels. This will determine the contributions of the principal peaks on the gates in distance of higher indices. If the contributions of several main peaks overlap, the composition law used is the general law

 In the timing diagram c) of FIG. 2, the following N determinations are represented by reading the buffer of the integrator in the decreasing direction of the remote gates.

As soon as there is detection, for the threshold k1 = &alpha; k0, of a principal peak
P of amplitude Ak high, the intermediate memory will receive on the 2 doors
P in the following distance, j = k - 1 to j = k - 2 '(decreasing direction) the attenuated value Sjkn = (Ak - x) n, in decibels. Thus, we will determine the contributions of the principal peaks index k on the gates in distance of lower indices. If the contributions of several main peaks overlap, the composition law used is the general law

On the time diagram d) of Figure 2, are represented, in the particular case where n = 2, the contributions of the main peaks, detected for the threshold k1 2, on each of the N gates in distance. These contributions are obtained by adding door-to-door squares of the contributions (ST) obtained in the increasing direction of the distance gates and the squares of the contributions (ST, obtained in decreasing direction) The composition law used is the general law
ST. n = (ST.) n1 + (ST) n with n = 2.

JJ ec To J2
The removal of the remote side lobes according to the method of the present invention is carried out by the device EL shown in FIG. 1 and described more precisely in FIG. 3. This device is used as a replacement for or in addition to the conventional system of FIG. threshold. However, it requires for its operation the use of a buffer MTI placed at the output of the detector-integrator. During the index recurrence period i, this memory will be intended to record the information provided by the integrator detector INT. During the next recurrence period, of index i + 1, this memory must be read in the increasing direction and then in the decreasing direction of the doors in distance while continuing to store the information provided by the detector-integrator. The principle of such a memory, using two memories in fact, is well known. In the case of a combined use of the threshold system S and the elimination device EL, a logic circuit L will, in addition, be necessary to carry out the synthesis of the information provided.

 During the reading, in the increasing direction, of the buffer MTI, each word read, corresponding to a gate in distance, is supplied to a first comparator CO, threshold kg, to a second comparator C1, threshold k1 = akO, a third comparator C2, threshold kg, forming part of the compensation chain and, via an attenuator A, on the first input of a switch AI.

This switcher AI controlled by the second comparator C1 provides on its output the attenuated value (Ak - x) in decibels, of the amplitude Ak of the principal peak detected by this second comparator, if the amplitude of this peak Ak is greater than the threshold k1 = ko of this comparator. It will supply on its output the value 0, provided on its second input, if the word considered, received from the buffer memory, is below the threshold k1 = akO.

The compensation chain comprises, in addition to the second comparator C1, the attenuator A and the switch AI, an ML device for raising the power n of the attenuated amplitude (Ak-x) in decibels, supplied by the switchman HAVE. This device may consist, for example, of a read-only memory. It provides, whenever a peak of large amplitude is detected, an attenuated amplitude, raised to the power n,
Sjn = (Ak - x) n, to a delay memory MR as well as to a "+" input of a first summer S1. This first summator is followed by an accumulator ACC whose output is connected to a second input "+" of this first summer S1. The delay memory MR introduces a delay of
P 2 doors in distance. From the reception by the first adder of an attenuated amplitude (Ak ~ x), the output of the accumulator, providing the first partial level of secondary lobes (ST will be charged to this value and each door in distance will reinject it into the summing up the output of that summator.
P 2 subsequent distance gates, no new large amplitude main peak is detected, this accumulator output will remain constant until the delay memory supplies its value Sk = (Ak - x) n on the input " - of summator E1 resetting to zero the outputs of the
P first summator and accumulator. If, during doors in
2 next distance, a new peak of large amplitude z is detected, then, at this moment, the first input "+" will receive from the read-only memory ML the value Sk, = (Ak, - x), the second input "+" will receive of the accumulator the value Sk = (Ak - x) n and the output of the first snmma- tor and that of the accumulator will provide a new value of the partial level of secondary lobes (ST = (Ak - x) n + - x) n corresponding to the sum of the powers n of the two contributions. As before, each of these two values will be deducted from this sum,
P 2 gates after their respective appearance, when these values will appear at the output of the delay memory MR. Each of the values (ST.) N supplied J1 by the accumulator during this reading of the buffer MTI in the increasing direction of the remote gates is stored in an intermediate memory MI to be used during the second reading of the MTI buffer performed in the decreasing direction of the doors in distance. During this second reading, the second partial lobe levels (STj) 2n corresponding to the peak contributions in the lower index gates are developed according to the same principle as the first (ST) corresponding to the contributions of the peaks in the gates. As they are developed, they are added term by term, by the calculator Cl,
n with the first contributions (STj) 1 stored in the intermediate memory MI. This calculator comprises at the output a device, of the read-only memory type for example, enabling it to output the
ith square root, or more generally the root n, of the square of the global level of secondary lobes STj2 = (STj) 2 + (STj) 2, respectively of its 1 2 power n.

 It is the second adder # 2, receiving from the buffer MTI the amplitudes of signal A. received in the j = 1 to N gate J in distance, which will provide the difference Aj = Aj - STj corresponding to the real amplitude, of the signal in the gate in distance of index j, freed of the contributions of the main peaks of large amplitude, or global level of secondary lobes ST. This real amplitude of signal A. will be compared in the comparator C2, of this chain of J compensation, at the manual threshold k0.

During the considered recurrence period of index i, the write phase is performed in the buffer MTI. The reading phase of this buffer memory in the increasing direction of the gates in distance, which makes it possible to determine the contributions (ST) of the peaks of great amplitude in the gates in distance of higher indices, can be entirely realized during the first half-period of recurrence next. The reading phase of this buffer in the decreasing direction of the doors in distance, which determines the seconds contributed

<tb> ions <SEP> (STj) 2, <SEP> global <SEP> levels <SEP>
<tb><SEP> ST. <SEP> = <SEP> (ST.) N <SEP> + <SEP> (ST.) Nn
<tb><SEP> J <SEP> J1 <SEP> J2
<tb> de-sidelobes and amplitudes A. of signal rid of the contributions of the main peaks J of large amplitude, can be fully achieved during the second half recurrence period following. The logic circuit L will therefore act during this second recursion half-period from the threshold crossing decisions D0j, D1j and D2j respectively provided by the three comparators CO, C1 and C2. This logic circuit L outputs a binary variable D such that
Dj = D1j + DojD2j which corresponds to the final decision.

 Although the present invention has been described in the context of a particular embodiment, it is clear that it is not limited to said example and is capable of modifications or variants without departing from its field.

Claims (8)

 A method for eliminating the secondary side lobes for Doppler radars with pulse compression, characterized in that for each of the N gates at a distance of a given recurrence period, it consists in evaluating the overall estimated level of secondary lobes STj in the distance index gate j and subtracting it from the amplitude A j of the signal measured in this remote gate, in order to deduce therefrom a real signal amplitude λ = A. - ST., unpolluted by the secondary JJJ lobes associated with the major amplitude peaks Ak, this actual amplitude λ being compared with the manual threshold k0 of the radar so as to determine if there is or there is no target in this gate at a distance.  2. A method of remote side lobe elimination according to claim 1, characterized in that the estimated overall level of secondary lobes ST. in the index distance gate j is calculated from the contributions Sjk of the main peaks of large amplitude Ak, detected in the gated distance gates, according to the general composition law

 where the contributions of the main peaks of large amplitude are limited to those of peaks present in the P remote gates  P P surrounding the door in considered distance, 2 front doors and 2 doors after, (with P even number).  3. Method of removing the remote side lobes according to claim 2, characterized in that the estimated overall level of secondary lobes ST. in the index distance gate j is decomposed J into a first contribution (ST) n of the principal peaks of large P 1 amplitude Ak detected in the 2 gates of indices greater than j and in a second contribution (ST. ) n of the main peaks of large amplitude Ak detected in the 2 gates of indices lower than j, the new law of contribution then writing:

 let STjn = (STj) n + (STj) n.  2 1  4. A method for eliminating remote side lobes according to any one of claims 1 to 3, characterized in that the detection of major peaks Ak large amplitude is performed by performing, on the integrator output (INTBune a priori detection of main peaks corresponding to useful targets whose level is greater than a threshold k1 = akg, this threshold k1 being chosen such that one has

where x is, in decibels, the level of sidelobes relative to the level of the principal peak, and, when a main peak of high amplitude Ak has been detected, by determining its contribution Sjk = Ak - x dB at level of sidelobes in the gate in distance of index j. <tb> <SEP> noise <tb> <SEP> #Prob. <SEP> <SEP> of <SEP> detection <SEP> of <SEP> targets <SEP> for <SEP> signal <SEP> # <SEP> 0 <SEP> dB # threshold <SEP> k0 <SEP> <tb> <SEP> noise <tb> #Prob. <SEP> <SEP> of <SEP> detection <SEP> of <SEP> targets <SEP> for <SEP> signal <SEP> # <SEP> x <SEP> dB # threshold <SEP> k1 <SEP> =  5. device for eliminating the remote side lobes according to the method according to any one of claims 1 to 4, characterized in that it comprises, at the input, a buffer memory (MTI) placed at the output of the integrator (INT) of the signal processing chain and allowing, simultaneously, to write the contents A. of the doors  J at a distance relating to the recurrence period of index i and reading the contents of these same doors in distance but relating to the recurrence period, index it, preceding and a device (C1, calculation chain A AC) for determining the estimated overall level of secondary lobes STj in the index distance gate j, from the signals Aj read in the buffer memory, a summator (z2) successively receiving buffer memory ( MTI), the signals A. read in the memory and, from the device for determining the  J secondary lobe levels, these STj levels and providing a comparator (C2), threshold kg, the actual signal amplitudes Âj = Aj - STj, this comparator providing a decision logic (L) decision D2j on the existence or non-existence of a target in the gate in distance of index j.  6. Distance lobe removal device according to claim 5, characterized in that a double read of the input buffer memory (MTI) is performed during the index recurrence period. first reading in the increasing direction of the gates in distance, realized during the first half-period of recurrence, making it possible to obtain the first contributions (STj) l of the main peaks of large amplitude on gates in distance of higher indices and a second reading, in the decreasing direction of the doors in distance, carried out during the second half-period of recurrence, making it possible to obtain the second contributions (ST.) n main peaks of large amplitude on gates in distance of lower indices, the estimated overall level of secondary lobes

 and the actual signal amplitude λ = Aj - STj.  7. distance lobe removal device according to one of claims 5 or 6, characterized in that the device for determining the overall level of secondary lobes ST. PP 2 consecutive distance gates, and reinjecting after 2 doors in distance the value, O or (au ~ x), which was then present on its input, to an input "-" of said summator (z1), this summator providing its output to an accumulator (ACC) which itself re-injects on a second input "+" of said adder (z1) its output and which is also connected to an intermediate memory (MI) for storing in memory during the first recurrence half-period the first contributions (STj) l and which will reinject, after a delay equal to this half-period of recurrence, the first contributions (STj) l in a calculator (CA) receiving on its other input the second contributions (STj) 2, provided by the accumulator (ACC), and developed during the second half recurrence period in the same way as the first contributions, this calculator (CA) outputting the overall level of secondary lobes STj in the door in distance of index j.  this delay memory (MR) introducing a delay corresponding to the reading of

 J in the index distance gate j comprises, at the input, an attenuator (A) and a first comparator (C1), threshold k1 = crkn, both receiving buffer (MTI) the contents A. distance gates read in the increasing direction then in the decreasing direction, a switch (AI), controlled by the output D1j of this first comparator (C1), receiving on one of its inputs the values A. read in the buffer memory ( MTI) and on its other input the value O and providing, depending on whether or not the peak level is below the threshold k1, either the value O dB or the value (Ak - x) dB, corresponding to the attenuated detected level, to a value of device (ML) of elevation to the power n, this device (ML) providing on a first input "+" of another summator (z1) and a delay memory (MR) its output  8. distance lobe removal device according to any one of claims 5 to 7, characterized in that it further comprises a comparator (CO), threshold kg, receiving the values A. read in the memory buffer (MTI) and a logic circuit (L) J receiving, during the second recursion half-period, the decision Doj elaborated by the threshold comparator (CO), receiving the decision D1j produced by the first comparator (C1) of threshold k1, receiving the decision D2j elaborated by the second comparator (C2) threshold kg and outputting for each of the remote doors, read decreasingly, the final decision Dj = D1j + D0jD2j associated with the door in distance d index j.