US3651412A - Receivers for pulsed signals - Google Patents

Abstract

In receivers of pulsed signals of a predetermined duration Tau , the useful signal is elaborated as a trigonometric function of the phase angle phi between the sum signal Sigma and difference signal of two samples, A, and B, of a received pulse at respective instants differing by t, t being smaller than Tau .

Description

United States Patent Poinsard et al.

RECEIVERS FOR PULSED SIGNALS Henri Poinsard; Marie-Jacques Jullien, both of Paris, France Thomson-CSF Apr. 22, 1970 Inventors:

Assignee:

Filed:

Appl. No.:

Foreign Application Priority Data 7 Apr. 28, 1969 France 69/13415 U.S. Cl ..325/476, 325/323, 325/324 Int. Cl ..H03lr 9/04, H0413 l/ 10 Field of Search ..325/323, 324, 476

[451 Mar. 21, 1972 L .l i V.

UNITED sures PATENTS I 3,423,682 1/1969 Cauchois ..32s/324 3,177,489 4/19'65 Saltzberg ..325/476 FOREIGN PATENTS OR APPLICATIONS 1,389,068 1/19es France ..;.....32s 6s Primary Examiner-Howard W. Britton Attorney-Edwin E. Greigg 57 ABSTRACT In receivers of pulsed signals of a predetermined duration 1, the useful signal is elaborated asa trigonometric function of the phase angle (/1 between the sum signal 2 and difference signal of two samples, A, and B, of a received pulse at respective instants differing by t, t being smaller than 1'.

9 Claims, 8Drawing Figures RECEIVERS FOR PULSED SIGNALS The present invention relates to receivers designed for the detection of pulse-type recurrent radio signals, in particular radar receivers, in which the effective signals are as picked up mixed with noise components and other parasitic signals.

In known art, after amplification, possible frequency conversion, filtering and detection, the signals are generally integrated that is to say that the signals, supplied by the video circuits at intervals of time spaced by T, T being the periodicity of recurrence of the effective pulse signals, are added up, this with a greater or lesser degree of accuracy.

This integrating operation has the effect of reducing the level of the majority of the parasitic signals in relation to that of the effective signals, the former generally not being recurrent or, if they are, generally not having the same recurrence frequency as the effective signals.

Among the signals obtained or integrated signals," there are, however, parasitic signals of substantial amplitude still left, in particular noise signals.

These parasitic signals constitute false pieces of information, the mean number of which per unit time must of course be limited if practical operation is to be preserved. This limitation is obtained by bottom clipping the signals produced at the output of the integrator.

All things being equal, the lower the clipping threshold, the higher the probability of a false alarm; however, the threshold should be regulated to the minimum value permissible in order to avoid excessive losses in terms of weak effective signals.

The mean amplitude of these false elements of information being an essentially random factor, it is necessary to regulate the threshold constantly or to provide automatic gain control.

In either case, the rapidity of the variations in this amplitude do not always make it possible to keep the probability of false alarm constant.

The invention relates to a radio receiver in which the false alarm incidence is statistically constant whatever the level of the signals picked up.

According to the invention, there is provided a radio receiver for the detection of recurrent pulse signals of predetermined duration 1-, comprising in series receiving means for receiving the recurrent signals, integrating means for integrating the received recurrent signals and clipping means for clipping the integrated signals, said receiver comprising further means inserted in series between said receiving means and said integrating means, for supplying a signal U which is a finite trigonometric function of the phase apglg 45 between twcL vectors A and B defined as the resultants A =2 j Kand 2 j Kwhere E is a vector representing the sum of the signals S (t) and S (t dt), S(t) and S (t+dt) representing the signals S received respectively at the instants t and t+dt, and d! being of the same order of magnitude as 1', where Kis a vector representing the difference between the sigrials S(t) and S(t+dt) and wherej A is a vector deduced from A by rotation through 11/2 in the trigometric sense.

In practice, simple functions will generally be adopted for U, either the cosine or, preferably, the sine function.

The signal U being independent of the absolute value of the amplitudes of the signals S(t) and S(t+dt), the bottom clipping threshold can be fixed.

The integrated signals corresponding to the random parasitic signals will be weaker in relation to the effective signals, the larger the number of periods of integration.

In practice, it is sufficient to effect integration over some few tens ofperiods.

In receivers of the kind in which the signals S(t) and S(t+dt) are presented at a sufficiently low intermediate frequency F the algebraical sums S(t) S(t-l-dt) and S(t) S(t+dt) can be effected at intermediate frequencies, either directly if f, is a multiple of l/dt, or else after phase shift of one of the signals; the signal U can then be obtained quite simply by applying to the two inputs of an amplitude phase detector, respectively the two sum signals after limitation thereof if the signal U cos is desired, one of the sum signals being in addition phaseshifted by 1r/2 if the signal U sin 45 is desired.

If the signals 8(2) and S(t+d!) are on the other hand available at video frequency, then it will be possible to respectively remodulate two auxiliary carrier waves of the same frequency and phase, either directly by the signals or by 8+8 and SS'. The signal U can then be obtained in the manner described hereinbefore, using an amplitude phase detector.

The signal U sin cancels out for S(!) S(t+d!). However, if the signal S(t) has, as a function of T, a symmetrical envelope with a maximum on the axis of symmetry, for example a substantially triangular envelope, the curve illustrating U as a function of t is an odd function of (tt,) cancelling out for t t,,, where t is the instantaneous value of t which defines the axis of symmetry. This curve, in the neighborhood of zero, has a shape similar to the error curve of a range tracking system the constant alarm probability receiver in accordance with the invention thus exhibits the interesting additional feature that it is utilizable, without any supplementary arrangements, for the range tracking of targets to this and the output signal from the integrator can be used.

The invention and its various applications will be better understood from a consideration of the ensuing description and by reference to the figures in which:

FIG. 1 illustrates an example of the shape of the signal detected by a receiver in accordance with the invention.

FIG. 2 is an explanatory vector diagram.

FIGS. 3 and 4 illustrate examples of signals produced in the receiver in accordance with the invention.

FIG. 5 is the general basic diagram of a receiver in accordance with the invention.

FIG. 6 is a diagram showing details of FIG. 5, in the context of one application of the invention.

FIG. 7 is a diagram showing details of FIG. 5, in the context of another application of the invention.

FIG. 8 is an example of application of the invention in an electromagnetic detection receiver.

By way of example, the invention will be described in the context of a receiver for an electromagnetic detection system which transmits substantially rectangular waveform pulses of width 1' at a constant recurrence frequency F,. The receiver must therefore detect from among all the signals received, those which are in fact echoes of the transmitted pulses. The targets returning these echoes are illuminated by the transmitter beam for a predetermined time T, which is a function of the beamwidth and also of the target width if the latter is not negligible. The N T F, echos from a target are processed in a filtering device, the internal characteristic of which more or less approaches that of the optimum filter. In a conventional receiver, integration during the time T produces a signal whose power is compared for each range quantum, with an adjustable threshold. An echo is defined as having been received, when this threshold is exceeded.

However, the mean background noise level varies during the course of operation and accidental noise effects of external origin are frequently superimposed upon it so that the probability of false alarms will vary since the threshold is fixed at least during the time of the recurrence.

In the receiver in accordance with the invention, the output signal from the optimum filter" is not directly integrated. An auxiliary signal U is produced which is a function of the relative values of the detected signals at instants differing by a predetermined time interval, but is independent of their absolute values. This is the signal which is integrated over N recurrences. If it stems from random parasitic sources, it is statistically zero (or at least very small): bottom clipping to the level of a fixed threshold is now justified, the mean noise being brought at any instant to a statistically constant level.

FIG. 1 illustrates as a function of the time interval t, the signal S(t) at the output of the optimum filter, this in the case where echoes of rectangular wave form pulses, of duration 1', are involved.

Depending upon whether the filter produces a video signal or an intermediate frequency signal, the signal corresponds to one of the envelopes ABC or ABC of width 2 'r or to the modulated carrier of envelope ABCB'.

Self-evidently, this representation is a theoretical one. In fact, the envelope is not strictly V-shaped but has a more rounded form, the essential as far as the present dissertation is concerned, being that it should be symmetrical, with zeros at the extremities and a maximum on the axis of symmetry.

The shape of this signal remains substantially the same for each of the N echos from one and the same target. Its amplitude is a function of that of the input signal to the filter.

A noise signal or more generally a parasitic signal, can randomly give rise during a recurrence, at the output of the filter, to a similar signal but it will not generally be reproduced during another recurrence.

First of all, the invention will be explained assuming that the processed signal is at video frequency. It will be shown hereinafter how the invention may equally be applied if the signal is an i.f. signal.

Let S S(t) be the value of the signal at the time t and S S (1+dt) its value at the time 1+ dt, dt being of the same order ofmagnitude as 1-.

If we call 2 the sum S+S, A the difference S-S', then the auxiliary signal U produced in accordance with the invention is a function of trigonometric f un c tions 9f the an le (1: between two vectorsA=2jA andB=2-jA ,where 2'andjAare two orthogonal (at right angles) vectors of amplitude Z and A respectively, as FIG. 2 shows.

The functions chosen by preference, are the functions sin (1) and cos 4:, these being obtained in a very simple manner as we will be seen hereinafter, and in particular the function sin d: which has special advantages.

FIGS. 3 and 4 represent cos 5 and sin (b as a function oftime in the case where dt -r, the signals S(t) having the form shown in FIG. 1. The signal cos is null for r/2.

For orders sake, should be pointed out that the curves representing sin dz and cos d) as a function of time, have been plotted taking, as origin of the abscissae, the instant at which the two values S and S are equal, that is to say the samples have been taken symmetrically in relation to the axis of symmetry of the signal.

In the case under consideration, sin 4) and cos d) are respectively expressed by and at a later point in this discussion the general expressions for sin 45 and cos 11: will be given.

Even from these simplified expressions, it can already be seen that:

the signal sin 4) is an algebraic signal in all cases, the signal cos only exhibiting two distinct polarities ifS and S do, in other words if two-polarity video working is used;

over several recurrences, the mean value of the signal sin 5 is zero for the parasitic signals, whatever they may be in other words, S, and S, of S and S for parasitic signal only are by definition identical so that their difference is zero;

over several recurrences, if the video signal is a two-polarity one, the mean value of cos 4: is zero for the thermal noise, this noise being en..irely decorrelated at the end of the time 1- however, it may not be zero for certain other parasitic signals.

The indicated approach, therefore, is to process the signal sin (b.

The effective signal will thus be obtained by the integration of the N signals corresponding to N successive recurrences for the echos the same signal will appear N times at the integrator input,-and the signal two-noise ratio is thus improved overall in the ratio V N.

In all cases, to the extent that the spectrum width of the noise is at least equal to the passband of the receiver, the integrated signal will in relation to the parasitic signals have a statistically constant mean value which makes it possible to fix the threshold once and for all.

lf,,on the other hand, the spectrum width of the parasitic signals is smaller than the passband of the receiver, N, being 5 the number of non-independent samples of parasitic signals (where N, is less than N), the ratio in effective signal to parasitic 7 signal after integration will be in the order of N N oth at the false alarm factor, for the same threshold, will be increased this can be remedied either by increasing the 10 time of illumination of the target in the ratio N/N or by using a transmitter-receiver system of rapidly varying frequency which means that the parasitic signals can be decorrelated from one recurrence to the next.

FIG. 5 illustrates the fundamental diagram of a receiver in accordance with the invention.

The signals picked up by the antenna A and possibly processed in the manner indicated in an input unit 51 or receiver proper, are applied to a device 52 which at the instant t dt produces at its output 521 the signals received at the instant t, and at its output 522 the signals received at the instant t dt thus, at the outputs 521 and 522 the respective signals S and S are obtained. These signals are combined in the operator 53 which produces at two outputs 531 and 532 The phase between these latter signals is measured in a phase detector 54 which produces the signal U as a function of this phase.

'At 55, the signals U are integrated, these appearing at the recurrence periodicity of the transmitter of the electromagnetic detection system in question the integrated signal is bottom clipped at 56 and at 57, the effective signal, of statistically constant false alarm factor, is obtained.

The above described diagram is a highly general one.

Depending upon the form in which the signals are supplied at the output 51, and thisessentially depends upon the nature of said device which does not form part of the invention, the following devices will be designed in different ways.

The receiver unit 51 may produce a video signal in which case the detailed diagram is that of FIG. 6 the unit 52 has two sampling devices 61, 62 which are set in operation a time interval dt one after the other by a clock 63, at the recurrence frequency of the system, and a delay 64, whose characteristic delay is dt, or r in the example described is mounted, in series with one of the sampling devices.

The unit 53 has two modulators 65 and 66 coupled to one and the same oscillator 67, the latter operating for, example, at 2 mc./sec., the modulators being followed by a conventional operator 68 comprising a combination of resistors and capaci- 5 tors and forming the sums S (l +j) S (l-j) and S (1j) S In the preferred case, where U sin (I: the unit 54 has two identical limiting amplifiers 609 and 610, one of them followed or preceded by a 1r/2 phase shifter 611, and an amplitude-phase detector 612 whose output supplies the integrator which latter can advantageously be designed in known fashion as a loop containing a two-input adder 613, one input of which is coupled to the output of the detector 612 whilst the output is coupled to one of the ends of the delay line 614 of characteristic delay T= l/Fr, the other end of the delay line being coupled both to the bottom clipper 56 and to the input of an amplifier 615 whose gain is close to but less than unity and whose output is coupled to the second input of the adder.

Alternatively, the receiver unit may produce an intermediate frequency signal in which case the device 52 may simply comprise, as FIG. 7 shows, a delay device 71 for bringing into time coincidence the signals received at the instants t and t dz, and, if need be, a device 72 for phase control, possibly with a feedback arrangement, the carrier frequency of the signals not generally being a'multiple of l/dt. In this case, the device 53 will simply comprise the operator 68, directly supplied with the output signals from the device 52.

If the signals are not brought exactly into phase, the expressions for sin 1 and cos I become:

LII

the denominators only being identical to S S if S and S are in phase these signals have the same advantages as the signals previously mentioned as far as the elimination of parasitic components is concerned.

The diagram of FIG. 7 is a very general diagram.

In FIG. 8, the receiver of a pulse-type electromagnetic detection system with fixed echo elimination and range channels, has been illustrated. This receiver is assumed to be of the coherent type, i.e., in which the echoes processed either stem from transmitted pulses, obtained by the chopping of one and the same carrier, or are detected using as a reference signal of each of them, a signal which is in phase with the carrier of the transmitted pulse. The receiver thus, in the conventional way, comprises beyond the circuit 51, which in this case produces intermediate or video frequency signals, a number M of parallel range channels the input gates 81.l,81.2,81.M ofwhich are successively opened during adjacent time intervals of duration 1, by opening pulses synchronized by the general synchronizing device of the system, 200.

Each of these channels comprises a fixed-echo rejection filter 83.1 83.2.... 83M.

The output signals from two successive channels being, offset by 1 the channels of successive outputs are grouped two by two and the output signal from the rejection filter for a channel thus constitutes both the signal S of said channel and the signal S of the next channel (M-l) operators 68,, such as the operator 68 of FIG. 7, in other words 68.1, 68.2.. 68 (Ml), have their inputs respectively coupled to the outputs of the filters 83.i and 83 (i+l respectively.

If the signals at the output of the device 51 are at intermediate frequency, (M-l) devices 52.], 52.2, 52 (Ml) identical to the device 52 of FIG. 7, are placed beyond each operator.

Each operator produces at two respectiveoutputs 8, and 9,, the signals A and B hereinbefore defined.

The outputs 8, are sequentially coupled by the electronic switch 100, controlled by the device 200, to the limiting amplifier 69 similarly, the outputs 9, are coupled sequentially in synchronism with the outputs 8,, by an electronic switch 101, to the limiting amplifier 620, which is identical to the one already described. The two amplifiers are coupled as before to the amplitude-phase detector 612, one of them through a 1r/2 phase shift element 611 if the function sin (1: has been chosen.

The output signal from the detector 612 is applied to the input of a switch 102 with (M-l) outputs, synchronous with the preceding ones, so that the integration ofthe signals U corresponding respectively to different range channels be effected correctly. To this end, each output of the switch 102 is coupled to an integrating device 55, preceded by a memory device 12,.

In order to determine whether or not there is an effective echo present, the signal U of each channel will be compared in 56 with the predetermined threshold, a further switch 103 being placed between 56 and the outputs of the integrators 55,.

In the case where the signals at the outputs of the rejection filters are detected, or correspond to Doppler frequencies which are very weak or differ two much from one another, for proper operation of the phase-detector circuit it will be necessary as in the case of FIG. 6, to sample the signals and to modulate an auxiliary carrier with these signals. Each range quantum will thus have to be sampled at time intervals in the order of magnitude of the time of decorrelation of the thermal noise as appearing at the outputs of the rejection filters, i.e., time intervals of less than T= HP, in the normal case where the passband of a rejection filter is small in relation to the recurrence frequency F By way of example, for T= 200/usec. and M 20, the measurement of each of the channels takes place during a time of less than lO/psec. and the frequency of the auxiliary wave can be made equal to 2 mc./sec.

The field of application of the invention is extremely wide since it covers all the cases of reception of recurrent pulse signals which are mixed with parasitic noise components, that is to say in particular it applies to numerous kinds of receivers used in aerial navigation and for space surveillance purposes.

It should be pointed out too that it does not merely apply to constant recurrent signal receivers in other words, the fundamental requirement is that the signals picked up from one and the same origin should be presented in accordance with a known law of repetition.

This is in particular the case of the kind of signal known by the name of bi-recurrent," currently utilized in electromagnetic detection systems in order to exclude blind zones."

Finally, it should be pointed out that in the case of a mobile range gate system for target tracking application, the invention at the same time makes it possible to effect acquisition of the target under good conditions since the false alarm ratio is constant and therefore the presence of echo" factor is optimized in order to effect tracking, it is then sufficient, once the presence of an echo has been indicated by a signal at the terminal 57, to process the signal sin 45 picked up at the input to the clipper 56, in the form of an error signal relating to the feedback loop controlling the displacement of the window, the latter being correctly positioned when sin d) 0.

Of course, the invention is not limited to the embodiments described and illustrated here purely by way of example. In particular, the delay dt could be other than 7.

What is claimed, is:

1. A radio receiver, for the detection of recurrent pulse signals S of predetermined duration 7, comprising in series receiving means for receiving the recurrent signals, integrating means for integrating the received recurrent signals and clipping means for clipping the integrated signals, said receiver comprising further means, inserted in series between said receiving means and said integrating means, said further means comprising a two input operator, means for applying to said two inputs the signals S S(t) and S' S(t dt), representing the signals received respectively at the instants t and t dt, and a! being of the same order of magnitude as r said operator having two outputs respectively supplying the signals A=(S+S) +j(S-S') and B (S-i-S') j (SS') two limiting amplifiers having respective inputs coupled respectively to the outputs of the operator, and an amplitude-phase detector having two inputs coupled respectively to the outputs of said amplifiers.

2. A receiver as claimed in claim 1 wherein, said receiver comprises detecting. means, said further means comprises two sampling devices for picking pairs of samples of signal S the two samples of a pair being taken at instants which are spaced by said time dt, said sampling devices having respective outputs, means for bringing into time coincidence the two samples of each pair, an auxiliary oscillator, having an output, two modulators having respective oscillation inputs coupled to said oscillator outputs, respective modulation inputs coupled respectively to said outputs of said sampling devices, and respective outputs coupled to the inputs of said operator.

3. A receiver as claimed in claim 1 wherein said further means comprises a direct channel and a delaying channel introducing a delay equal to said time dt said channels having respective inputs coupled to said integrating means and respective outputs respectively coupled to two inputs of the operator.

4. A receiver as claimed in claim 3 wherein one of the said channels comprises a phase-shifting device.

5. A receiver as claimed in claim 4 wherein a phase feedback devices is coupled to the said channels for controlling said phase shift element.

6. A receiver as claimed in claim 6 for a pulse radar system transmitting recurrent pulses of duration 1' having a substantially rectangular waveform, said receiver comprising a filter device which converts the signals of width 1' into signals with a substantially Vhshaped envelope of width 21-1- M (M being an integer greater than 1) range gate channels, numbered 1 to M successively opened during adjacent time intervals equal to -r each channel including a fixed echo rejection filter having an output, said receiver further comprising (M-l) operators, numbered 1 to M-l an operator numbered i having two inputs respectively coupled to the filter outputs of the channels numbered i and i l said operators respectively having first outputs supplying said signal A and second outputs supplying said signal B a first synchronized switch successively coupling said first outputs of said operators to one of said limiting amplifiers, a second switch, operating in synchronism with said first switch for successively coupling said second outputs of said operators to the other limiting amplifier; integrator circuits numbered 1 to M-l the-integrator numbered 1' being associated with the channels of order i and i+ l a third switch, synchronously operated with the former two, coupling the output of said phase-detector successively to the (M-l) integrators; a fourth switch successively coupling the outputs of the phase-detector successively to the (M-l integrators; and a fifth switch successively coupling the outputs of the integrators to the clipping means.

7. A receiver as claimed in claim 6, designed to cooperate with a pulse-type electromagnetic detection system comprising a sliding range gate having a control input; said receiver comprising means for elaborating signal U sin #1 and for feeding said signal U to said control input.

8. A receiver as claimed in claim 6, in which the output signals of the rejection filters are if. signals, further comprising, between the outputs of two successive rejection filters of order i and i l and the inputs of the operator of order i respective (M-l delaying phase shifting devices.

9. A receiver as claimed in claim 8, in which the signals at the output of the rejection filters are video frequency signals, further comprising sampling devices respectively coupled to the output of said rejection filter.

k II

Claims (9)

1. A radio receiver, for the detection of recurrent pulse signals S of predetermined duration Tau , comprising in series receiving means for receiving the recurrent signals, integrating means for integrating the received recurrent signals and clipping means for clipping the integrated signals, said receiver comprising further means, inserted in series between said receiving means and said integrating means, said further means comprising a two input operator, means for applying to said two inputs the signals S S(t) and S'' S(t + dt), representing the signals received respectively at the instants t and t + dt, and dt being of the same order of magnitude as Tau , said operator having two outputs respectively supplying the signals A (S+S'') + j(S-S'') and B (S+S'') - j (S-S'') , two limiting amplifiers having respective inputs coupled respectively to the outputs of the operator, and an amplitude-phase detector having two inputs coupled respectively to the outputs of said amplifiers.

2. A receiver as claimed in claim 1 , wherein, said receiver comprises detecting means, said further means comprises two sampling devices for picking pairs of samples of signal S , the two samples of a pair being taken at instants which are spaced by said time dt, said sampling devices having respective outputs, means for bringing into time coincidence the two samples of each pair, an auxiliary oscillator, having an output, two modulators having respective oscillation inputs coupled to said oscillator outputs, respective modulation inputs coupled respectively to said outputs of said sampling devices, and respective outputs coupled to the inputs of said operator.

3. A receiver as claimed in claim 1 , wherein said further means comprises a direct channel and a delaying channel introducing a delay equal to said time dt , said channels having respective inputs coupled to said integrating means and respective outputs respectively coupled to two inputs of the operator.

4. A receiver as claimed in claim 3 , wherein one of the said channels comprises a phase-shifting device.

5. A receiver as claimed in claim 4 , wherein a phase feedback devices is coupled to the said channels for controlling said phase shift element.

6. A receiver as claimed in claim 6 for a pulse radar system transmitting recurrent pulses of duration Tau having a substantially rectangular waveform, said receiver comprising a filter device which converts the signals of width Tau into signals with a substantially V-shaped envelope of width 2 Tau Tau , M (M being an integer greater than 1) range gate channels, numbered 1 to M , successively opened during adjacent time intervals equal to Tau , each channel including a fixed echo rejection filter having an output, said receiver further comprising (M-1) operators, numbered 1 to M-1 , an operator numbered i havIng two inputs respectively coupled to the filter outputs of the channels numbered i and i + 1 , said operators respectively having first outputs supplying said signal A and second outputs supplying said signal B , a first synchronized switch successively coupling said first outputs of said operators to one of said limiting amplifiers, a second switch, operating in synchronism with said first switch for successively coupling said second outputs of said operators to the other limiting amplifier; integrator circuits numbered 1 to M-1, the integrator numbered i being associated with the channels of order i and i + 1 ; a third switch, synchronously operated with the former two, coupling the output of said phase-detector successively to the (M-1) integrators; a fourth switch successively coupling the outputs of the phase-detector successively to the (M-1) integrators; and a fifth switch successively coupling the outputs of the integrators to the clipping means.

7. A receiver as claimed in claim 6, designed to cooperate with a pulse-type electromagnetic detection system comprising a sliding range gate having a control input; said receiver comprising means for elaborating signal U sin phi and for feeding said signal U to said control input.

8. A receiver as claimed in claim 6, in which the output signals of the rejection filters are i.f. signals, further comprising, between the outputs of two successive rejection filters of order i and i + 1 , and the inputs of the operator of order i , respective (M-1) delaying phase shifting devices.

9. A receiver as claimed in claim 8, in which the signals at the output of the rejection filters are video frequency signals, further comprising sampling devices respectively coupled to the output of said rejection filter.

Images