CA1177572A - Active detection system using simultaneous multiple transmissions - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

A system for detection by an echo containing transmission and reception sources for detection by acoustic waves in particular. A circular base contains N columns T1, T2,..Ti,..TN of electroacoustic trans-ducers. To each column, such as Ti, corresponds a transmission at the frequency fi. A switch makes it possible to change from transmission to reception. The signals received are applied to a channel forming device. For each channel, such as Vk, the signals are filtered at frequencies f1, f2..fN, which are applied to time offset circuits taking into account the base geometry and the channel direction. The signals obtained, such as di.k, are applied to a display device which supplies the range, direction and speed of the targets detected. This is particularly applicable to underwater surveillance by sonar or radar.

Description

A~ ACTIVE DETECTION SYsTE~ USING SIMULTANEOUS ~LTIP~E
TRANSMISSIONS

B~CKG~o~ND OF T~E INVE~TION

The purpose of the Present invention is an increase in the resolution of active detection systems whose transmissions may be either electroma~netic, s~lch as radars, or acoustic~ such as sonars, In these systems which are intended either for target direc-tion or for extraction and identification or classification of certain parameters of possible targets or for image forming, the mode used, for sonar in particular, to illuminate the angular field ~o be o~served is that of a single, coherent transmission, which covers the whole angular-field to be observed, of duration T and bandwidth b emitted round a carrier frequency f , while for reception, a set sf angular channels is formed in parallel to cover this field.
The angular resolution of these systems is virtually that of ~5 the recePtion antenna and it is known that its dimensions add a limiting factor to this resolution.
It is known to increase the spatial resolution of systems in the sonar field with a linear acoustic base which uses, for transmission, a so-called interferometric mode that consists in transmitting two signals at the same frequency simultaneously from two transducers at the ends of the base.
The result of this transmission i9 the formation in space of zones that are sounded and unsounded alternately and, when convoluted with preformed channels on reception,Igive directivity improved by a coefficient of 2, the existence of ehe unsounded zones being the disadvantage of the method.
This can be corrected by two separate methods. Either a second simultaneous transmission is used, which is offset angular-ly with respect to the preceding one and which doubles the sweep time or two simultaneous transmissions are used, which are ~ 177572 separate in frequency, one having the characteristics of the first transmission ab~ve and the other those Qf the otber transmis-sion, but then the useful band is doubled.
Another example of i~provement in the directivity can be found in American patent N 4 119 940 - Keating, which also uses N frequencies for transmission. This invention, which separates well the signals produced at the different frequencies but causes them later to lose their identity related to their different frequencies and does not do any space-time processing on them, is applicable neither for base-object relative movements for non-planar base.
SUMMARY OF TNE INVENTION
The system in accordance with the invention corrects these disadvantages. FQr the angular resolution, the system proposed is identical to a system which forms channels both on transmission and reception, i.e it gives an angular resolution double that of a system which only forms them on reception.
For resolution in range and "Doppler speed", the system proposed is identical to a classical active system which transmits a signal of length T and band B on each recurrence, i.e one range resolution of B and one Doppler fre~uency resolution of T in which c is the wave speed.
For the data rate, the system proposed is identical to one transmitting a single signal of duration T and band B on each recurrence and forming narrow channels on rec~ption only.
- Also, the system has the advantage that it can be applied to certain equipments already existing.
These equipments are those whose transmission source are fed by modular transmitters and whose channel formations are wide band, The invention can be applied to them by coding the pilot signals of the transmitters at low level and adding, after the formation of the reception channels already existing, signal decoding in these channels followed by temporal offsetting of the signals decoded.
In sh~rt, it is a system for transmission and target detection by echo reception which contains :
~ sources El,..., EM forming a transmission array of depth d, i ,~

~ :~7~S7~
which tr~nsmits simult~neously wave~ ~dulated by signals coded Cl,..., CM, the codes being different and separable, reception sources Hl,... 9 ~N~ forming a reception array which receives signals in an overall band W, means for processing the signals received, means for using the signals processed, wherein the processing means contain, operating together ;
- means for forming channels, called reception channels, in I ~ 2N with directional functions D D
means for decoding the signals received adapted to codes Cl,...CM
in each of the so-called reception channels, means for forming, on reception, channels, called transmission channels, with directions and direc~onal functions roughly identical to the corresponding reception channels.
Other characteristics and advantages will appear from the description which follows.
BRIEF DESCRIPTION OF THE DRAWINGS
.
Fig. 1, the detection system ~ith simultaneous transmission in accordance with the invention.
Fig, 2, the general block diagram for a sonar with a circular base, Fig. 3, the processing diagram for a sonar with a circular base.
Fig. 4, details of circuits in fig.3.
Fig, 5, a representation of codes used for transmissi~n.
Fig. 6, the schematic diagram for a variant of the invention.
Fig, 7, the appearance of the signals for the variant in fig,6, DESCRIPTION OF THE PREFERRED EMBODI~ENT
Assume a set of sources Sl,.. Sp of which M are able to have the transmission functions Ei, (] < i < ~) and N the reception functions F , (I < j c N). The position of these sources are such that channels can be formed both on transmission and reception.
The distribution of these sources may be random provided that their coordinates are known. However, the most commonly used are tri-rectangular, spherical, cylindrical, circular, conforming, etc, distributions, The distribution of the energy in space or directivity D, which results from the channel formation on ~, 1 1~75~2 transmission ~nd Feception is gene~ally expressed by m~ans of two ~ngula~ parametera D (~9 ~ in ~ horizontal and ~ in a vertical plane, When these conditions have been stated, the purpose of the invention is an improVement in the spatial resolution of the system by superimposing the directional func~ion on transmission DE (~ and the directional function on reception DH (~
along roughly the same direction (~m~ ~n) with the condition that DE (~ ~) ~ DH (~ ~) no matter what the values of ~ and ~, the result being an overall directivity D (~, ~)2 improved by a factor of 2.
It is known that the directivity may be improved by superim-posing the cha~nels formed on transmission and reception. However, the processing of the signals in accordance with the invention enables the disadvantages of known practice to be avoided. The system proposed is based on the combination of the transmission and reception functions.
In the system in accordance with the invention, M signals, coded Cl,..., C~,... CM~ which are separable in band b and of the same duration T, are transmitted simultaneously by means of M sources, El,... Ei,... EM, whose space coordinates are known, The main types of t~ansmission possible are, for example (fig.5) ;
I - M signals of pure frequency fl~ ~ fi ~ f~ of duration T separated in frequency by T = b,

2 ~ M signals of the same centre frequency f coded in the same band b with M orthogonal codes of duration T in which bT > M,

3 - M signals coded in the same band b with the same code 30 of duration T in which bT > I and with different centre frequencies separated in frequency by the distance b,

4 - M signals coded in the same band b with Q orthogonal codes (Q < M) of duration T in which bT > Q and P different centre f~equencies (PQ = M) separated in frequency by the distance b.
- 35 ~s f~r as receptign is concerned, this consists in receiving at N transducers Hl,... HNg the echoes from these transmissions, each transducer receiving all the echoes during the recurrence I :~7~5~72 which come f~om the w~ve~ t~nsmitted by ~11 the transmitter~, ~I to EM, these signals being p~cessed in three fitages ;
I~ F~rmation in real time, i.e, in parallel, of all the reception channels Vk with directions (~m~ ~n) from these transducers Hl,.., ~ , these channels having a certain directivity D~
2) sorting by matched filtering in each rection t~ ~ ~D) thus obtained, of echoes coming from the various transmissions El to EM, 3) formation of directional transmission channels with a directivity DE (~ ~) roughly identical to that of the rec~ption channels and in the same directions by temporal offset which consists in superimposing, in each of these directions (~ ~ ~n)~ all the echoes thus sorted by suitable delay or phase means and summing them in amplitude and phase for each possible range slot.
Fig. I shows these various stages.
I represents an array, which may be conforming, i.e has the shape of its carrier (the stem of a ship for a sonar), consisting of M transmission sources El to EM, which receive from M generators (not shown) their electric signals coded Cl to C~ in accordance with one of the codings mentioned above and N reception sources Hl to HN.
In ci~cuits 2 the first signal processing operation takes place which consists in forming in parallel (or in series if the time taken to do it is less than the reciprocal of the overall band) all the reception beam channels in the directions (~
required for the system, This formation of channels in the band W = B (transmission) ~ D ~in which D is the maximum variation in frequency due to the Doppler effect) can be obtained - either by means of devices giving suitable delays in the case in which the de?th of the array, i.e the distance covered by the waves tQ reach all the array transducers in the direction ~ ), which is the most inclined of the channels formed, is greater than W' - o~ by means of deyices giving suitable phase shifts in the case in which this depth is less than W for each of the signals coming from the transducers Hl,... HN, this operation I :~ 775~2 being ca~ied out ~ different nu~be~ of times p corresponding to p separate directions ~
At the output of circuit 2 there are then p channels ormed, Vl,.., Vk,.. V , each containing all the echoes coming from all the transmitters El to EM.
The signals from each of these channels, such as Vk~ are applied to a c;rcuit such as 3.k in which is carried out the second operation of signal processing which consists in sorting the echoes, in each direction of reception (~m~ ~n) previously obtained, in accordance with the points in the transmission space which are at the origin of their production. This is made possible by the coding of the transmiss}on whic~ identiy the ~ransmitters El,,,E~. This sorting is done by means of filters matched to the t~ansmîssion codes, the Doppler effect of the targets being taken into account. A simple example of matched filter can be given in the principle of coding I given abovè (pure frequencies of length T
separated in frequency by T) The sorting operation after each of the reception channels Vk (~, ~) then consists in a spectrum analysis of resolution T in an overall band W = M x T ~ D, in which D takes into account all the possible relative movements of the objects to be observed. This sorting of signals at different frequencies can be obtained.
- either analogically by the use of M ~ DT = R simple band pass filters in parallel separated in frequency by T (simultaneous data) qr by the use of Fourier transforms with dispersive delay lines. This type of processing is described in "Use of dispersive delay line for signal processing in underwater acoustics" by P.
Tournois and J. Bertheas, JASA 1969, pp. 517-53}.
- o~ digitally by processing such as a fast Fourier transform (F.F,T,~ with a resolution of T.
There are R different outputs, as many as there are channels analyzed in the whole band (figure ] only shows M7 i.e the case without Doppler), at the output of circuits 3.l, 3.2,...3.p and `fo~ each reception channel.
Another example of an matched filter can also be given in the second coding in which all the signals are transmitted at the same central frequency but with separable codes Cl,... Ci,...

~:~7~5~2 CM called orth~gon~l codes. The role of this filter consi~t8 in co~paring the different codes ~eceived with the code Ci it has in its memory. The comparison supplies a signal, called a correlation signal, which is a maximum when the signal received S contains the code Ci among all the others and a m;nimum when it does not.
This matched filter is a correlator. ~This processing is described in "Digital communications" by S.W. Colomb (Prentice Hall) with examples of orthogonal codes).
In this second example, the sorting of the signals received will be carried out after each of the reception channels by R
correhators, as many as there are different codes in the overall band W = b + D of signals received.
The third coding corresponds to a double coding : to M
~5 frequencies aud inside each of these frequencies.
The matched filter is obtained by spectrum analysis followed by a single filteradjusted to the specific code. If this code is, for example, that of a transmission modulated linearly in frequen-cy, the correlator will be a filter with a matched delay law (this processing is described in the article already mentioned "Use of Dispersive Delay Line").
The fourth coding is a mixture of the preceding codings. Its interest lies in the reduction of the band required by the third coding type for example. For example, this band can be divided by two by placing two orthogonal codes at each of the frequencies fl to fM~2, which are modulated linearly in frequency one with increasing frequencies and the other with decreasing ones.
The third reception s~gnal processing operation is carried out by circuits 4.1,.,.4.k,.,. 4.p and consists in the algebraic reconstruction of the transmission channels. This algebraic Feconstruction is obtained, as for the forming of the reception ch~nnels discussed above, by the introduction of delays ~1 or different phases Ui = ~O T not into the signals coming from reception sources Hl,.,. HN but into the signals previously sorted and corresponding roughly to the Doppler effect, from sources El,..
EM and this for each direction (~ , ~ ) of the reception channels.
The processing in circuits 4.1,... 4.p corresponds physically ~ :L7'~72 to two effects ;
~ an effec~ of 'ldi~ectivity on transmission" since the offset is only valid for a given direction, - - a temporal effect of "pulse compression" whose explanation can be given more easily if the signals transmitted for the first coding are considered. The enveloppe of the superimposition or "offset" and the algebraic sum of N signals of separate pure frequencies, distant I from one another in a band B = N~ is absolutely identical to a short signal of SlnBT-BT of len~th -, i.e N times shorter than the original signal.
An object, which is receding or approaching with a radial speed V shifts in phase the transmission frequency fi by the quantity ~ - x fi = D as a first approximation if V c.
More generally, this frequency shift will result in a disturbance to the codes received which may lead to a zero result for convolution with the matched filters.
In the case of the f;rst coding, the frequency shift causes an apparent angular rotation of the transmission sources and the direction of the transmission beams formed may no longer corres-pond to the direction of the reception beams formed. This is theproblem of sensitivity to the Doppler effect.
The solution consists in using a transmission code type, a t~ansmission-reception source geometry and a frequency distribution which are less sensitive to the Doppler effect and in using a larger number of matched filters, in the limit one per Doppler resolution step.
Fig. 2 shows the block diag~am of a particular production of the invention applied to a so-called panoramic sonar, because it is able to cover the whole horizon, in which it is desired to increase the directivity of the channels formed by a factor of 2 with respect to prior practice.
A circular base 12 contains N columns of transducers such as Ti which are connecLed to transmission-reception switching ci~cuits 7. The signals received are applied to beam forming circuits 8. There are the elemen~s of figure I matched to the fi~st coding case. Each of the chalmels for~ed in circuit 8 is analyzed in frequency by circuits 9.1,... 9.k,...g.2N and the frequency sa~ples are then off~et in time and 6uperimposed in tran~missi~n ch~nnel for~ing circuit~ 10,1,.., 10~2N which supply the successive Doppler samples for a giyen (transmission receptiOn) channel displayed by the operating system l1.
The shape of the signals obtained at the output of circuits 8, 9.1,... 9.2N, ... 10~1,... ~0.2N is shown in figure 2.
At the ou~put of channel forming circuit 8, S~,... Sk,...
52N~ the signals of the channels formed Vl (~),.. Vk (Q),.. V2N(~) are obtained. It ;s assumed that the target was in the direction corresponding to channel Vk (~) and that the other signals only represented noise.
At the output of spectrum analysis circuit 9.k there are R
frequency analysis channels only M of which contain a useful signal, the others ~ - M also represent noise. The M useful signals are offset in time in accordance with the geometry of the base 12 by times such as T ~ .
At the output of transmission channel forming circuit 10.k, the preceding M useful signals are offset in time to compensate for delays such as ~i in the direction of channel Vk(~).
Because of the nature of coding in frequency, the duration ~f the si~nals obtained in one of the Doppler channels dj k at the output of circuit 10.k is compressed in the ratio BT
and its amplitude is increased by a factor of tBT)0'5 dj k correspond6 to the speed of the target detected in channel Vk(~).
Fig, 3 shows the detailed block diagram of the preferred embodiment for the assembly in fig.2. The cylindrical acoustic base ]2 consists of N identical columns. The columns tr,ansmit (functions E1,... Ei,...EN) and receive (functions Hl,~ Hi, . HN) alternately~
All the transmissions are simultaneous, of duration T and band b = T~ the pure frequencies f~ fi~ - fN generated by genera~ors 14 being applied respectiYely to the columns. This is then the transmission mode for the first coding.
Switches 7 pass the transmission signals to the respective columns and the reception signals H19. . Hi,... HN to the reception channels, one per column, formed by circuits 81, 82, 83 and 84.

I ~ 77~7~
Ci~cuit 8I represents ~ preamplifieF a~d level regulator. Circuit 82 is a band pass filter of Width B ~ D, Ci~cuit 83 represent~
~ frequency changer intended to bring the centre frequency fo =_I _f2 to around zero frequency, i.e transfer the réception r 2 spectrum to the so~called "b~se" band where the real and imaginary par~s of the signal are transmittçd separately. These two compo-nents, out of pbase by ~/2, are nPcessary to keep both the ampli-tude and phase data of the signals received. This operation, which is not required when the reception channels have been formed in circuit 86, is carried out to obtain the complex product.
A low pass filter 84 filters the intermodulation products Qbtained beyond the band ~ . Sampler-multiplexer circuit 85 operates at the clock frequency FE = 1.25 (B + D), which is higher than the frequency determined by Shannon's theorem. At the output of 85 the samples of the signals moved from the succes-si~e columns are fed to the charge coupled device (C.C.D.) forming reception channels by interpolation.
Interpolation also enables 2N channels to be formed, i.e t~wice as many as the number of columns. Because of the increase in angular resolution provided by the present invention, the num-ber of channels N of prior art is here doubled, i.e it rises ~o 2N.
At the output of devices 86 the successive analog samples of the channels formed corresponding to the directions ~ Qk' ~2N on successive turns exploring the horiæon are to be found.
Circuit 87 is an analog-digital converter in which the pre-ceding analog samples are converted to digital samples in 8 bits.
Circuit 88, a demultiplexer, is there to replace in parallel the signals from the 2N channels which we~e previously formed in series. Assembly 91 is a digital memory of the random access (RA~) type consisting of 2N memor~es such as 91.k, where I <k <2N, intended to retain a duration T' = T ~ - for the successive samples in the same channel ~k~ - in which d is the array depth alrPady defined, Each memory 91.k is double, (91.k)R and (91.k) which are filled simultaneously by demultiplexers 88. The number of memory elements in each of them is FE.T'. This duration T', the length of the signal recorded, enables a fast Fourier transform (F.F.T.) to be produced in computer 92 with a frequency ~ 17'7~'72 resolution of T. The ~eason for the presence of multiplier circuit 61 will be better understood by refe~ence tQ figu~e 4 uhich giVes details of the blocks such as 9.k and 1~.k, which form assembly 13 in figure 2, in which the operations of freqtlency analysi~
and time offset respectively of the transmissions are carried out.
Circuit 91 is the memory already mentioned. The fast Fourier transform device 92 calculates the successive samples of the spectrum of the frequencies contained in the section T of the signal received by the reception channel Vk ~). These successive frequencies are converted to analog signals in analog-digital converter 101 and passed to circuit 102 forming transmission chan-nels by temporal offset of the frequency signals.
This transmission channel forming can be done, for example, b~ a charge coupled device since the offset depends on the array geo~etry as for the reception channel forming. As has been seen in this production, which operates in a narrow band, the forming of channels was really a simple phase offset. Hence, this forming of transmission channels in the frequency field is a simple frequency convolution between the spectrum gk (~) received from the direction ~k and the spectrum hk (~)~ a complex conjugate of the spectrum and all the signals emitted at infinity in the medium and in the same direction. The known relationship (Plancherel's theore~) makes it possible to pass from a convolntion in the frequency space to a simple product in the time space. If then Gk (t) and ~ (t) are the Fourier transforms of the respective functions gk (~ and hk (~)~ then the Fourier transform of (Gk (t).Hk (t))is gk (~) * hk (~) a relation which shows that the conyolution operation carried out in circuit 102 after the Fourier transform carried out by computer 92 can be replaced by a simple time multiplication, carried out term by term before the Fourier transform.
Circuit 61 in figure 3 consists then of two digital memories for each channel Vk (Q), of the programmable read only memory type ~P~OM 611,k and 612.k and mul~iplier circuits S13.k and 614.k.
Memories 611.k and 612.k contain respectively the real and imaginary parts of the impulse response in the matched filter to the various codes superimposed and transmitted in the direction ~k. ~his ~ :~7~5~2 reco~ded tempo~al signal is merely the physical signal S(~k,t) returned in time~ receiyed at a point distant from acoustic base 12 in the direction ~k and converted to the base band. Hence, it only depends on :
- the code transmitted by each of the columns, - the geometry of the columns with respect to direction ~k~
a geometry which introduced the various phase shifts in this direction, - possibly, a window for weighting the function recorded.
The nu~ber of cells in memories 611.k and 612.k is the same as that in memories 91. The T.FE samples in memories 91 are brought out and multiplied term by term by the T.FE samples in memories-611.k and 612.k in multipliers 613.k and 614.k of the real and imaginary parts.
In computers 92.k are the operators for the fast Fourier transform which operate on T~FE points at the analysis recurrence of N.
As many F.F.T. computers are shown in figure 3 as there are reception channels. However, the computing speed of these operators enables the computing work of several reception channels, four for example, to be assigned in sequence to a single operator.
The samples obtained at the output of computers 92 represent the re~l and imaginary parts of the frequency analysis of the signals in a given channel after its double formation. They are therefore the Doppler channels in series obtained at the frequency step of T. There are D.T of them.
In computing circuit 103 the square of the modules in the real and imaginary parts, R and I, of these Doppler frequency samples is calculated.
The device using this data in a display receives the 2N
~eception~transmission channels with the data in each channel concerning target range and speed.
The display uses a colour tube with the îmage (~, range~
in mode B. The Doppler data are applied to the chrominance signal.
~n a preferred production there are the following characte-~istics ~

1 .lL~757~

Acoustic base diameter ~ - 2.5 m Number of columns N = 32 Transmission length T = 0~3 sec F~equencies i N3500 - 53 Hz Transmission band B = T a 106 Hz DoppleF band for v = 30m/sec D = ~ 140 Hz i.e. 280~z Ov~rall band W = B ~ D = 386 Hz Sampling frequency per column FE = 426 Hz Distance covered ~y the wave c between two samplings F = 3 50 m (compare with Number of memory cells (91.k) T.FE = 828 elements F~F.T. analysis recurrence T
tiMe ~ = 9.4 msec Total number of F,F.T calcula- 2.N
tions per second T = 6827 F.F.T./sec Number of elementary operations T.FE
in an F.F.T. 2 = log2 (T.FE) = 448 Total number of elemen~ary operations in an F.F.T 6827 x 448 = 3,058,496/sec Num,ber of ~.F.T. computer cir-cuits 92 ~ith a 1.3~sec cycle Number of Doppler channels DT = 84 In a variant of the invention the two operations carried out in circuits such as 3.k and 4.k respectively for matched filte,ring for the transmissions matched and transmission channel fo~ming can be replaced by a single operation carried out by a device for conyoluting the signal received in each channel with the copy of the sun of the signals emitted at infinity in the direction corresponding to this channel and inverted in time.
Fig. 6 shows the general block diagram of this variant of the invention which replaces all the processing carried out by circuits 9.1,... 9.2N and 10.1,.. 10.2N.

13 , Memories 15,R and 15~I are add~essable ~emories of the RAM
type in which the ~eal and imaginary signals receiyed in the channels formed are w~itten in columns and read in lines, one line co~esponding to one channel~ ~ddressing of the signals on

5 - reading and l~iting i done by circuit 16. Writing is done at the same frequency as the sample output frequency of the channel for-ming circuit 86.
The duration of the signal recorded in each line is T' = T + ~t Reading of these lines is carried out at a higher frequency H1 so that the signal of length T' and band T~ now occupies a band T-'-The signals of the lines corresponding to the channels formedundergo a digital/analog con~ersion in circuits 17.R and 17.I and are then added to a carrier F1 by complex modulator 18 which re-ceives the components sin (2~F1.t) and cos (2~F1.t). The signal V1 (t) coming from this modulator is passed to a convolution analog device 19 which is preferably an elastic wave device as described in the Proceedings of the International Seminar on Component Performance and System Application of SAW Devices, 20 E.C.S Paige, pp. 167 - 180, 1973 (an I.E.E. publication) or in the I.E.E.E. Proceedings, P. Defranould and C. Maerfeld, p. 748, May 1976.
The centre frequency of ~his convoluter is Fi and its band F' This convoluter also receives the convolution copy V2 (t) defined above and corresponding to the same direction. It supplies a signal C(t) which is demultiplexed by circuit 20 at whose output the various channels formed in parallel are obtained, each of the channels containing in series the three samples in each Dop-pler channel. These signals are used by display dlvice 11 in Fig.3.
Signal V2~t) is obtained from memories 21.R;and 21.I which conta;n, in digital form, the copies, converted to the base band, of the signals emitted at infinity and inverted in time. These signals are stored in line, each line corresponding to a given direction.
This memory is of the programmable, read only memory type (PROM~.

~ :~7~72 Thes~e d~ta a~e ~ead in line$~, each line being read R times, R being the number of Doppler channels, Addressing is done by circuit 22 at the r~te of ~2~
The data 2ead in the two memories 21,R and 21.I are conver-ted into analog signals by circuits 22.R and 22.I and a~ded toa carrier at frequency F2 by circuit 239 which is identical to circuit 18.
Carrier frequency F2 and clock frequency H2 are generated by generators 24 and 25 respectively which are of the voltage-control-led oscillator (V.CØ) type, the shape of this voltage beingshown in Fig.7.
The whole device shown in Fig. 6 operates as follows. To each line read in memories 15.R and 15.I and fed to convoluter 19 with carrier frequency F1 at the rate H1, both of which are fixed, corresponds the homologous line in memories 21.R and 21.I
fed to convoluter 19 with carrier frequency F2 at the rate H2, both of which are variable, this frequency and rate being changed R times. In each memory the homologous lines are then read R
times, the first with fixed frequency and rate and the second with variable frequency and rate.
The amplitude of the variation in this Erequency and rate around F2 and H2 are respectively : ~F2 = K. . max and ~H2 = H1 x ~a~ .. This variation is obtained by R eqllal jumps and V is the maximum speed of the target. For preEerence Kfo = F1-The variation ~F2 and ~H2 during successive readings of the same line give copies of the transmission signals modified by a Doppler effect. This gives a signal in the Doppler channel corresponding to the target speed.
Detection systems have thus been described which, with an array of given dimensions, enable an improvement in the angular resolution to be obtained.
Evidently, this description has only been given as an indication for Sonar. The present invention is applicable to a Radar whose rotating antenna has been replaced by fixed sources distributed on a wave. In this case, the system enables an ~ ~7~7~

improyement t~ be obt~ined, nQt in the angula~r ~r~esoluti~n but in the d~ta r~te as a function of the numbe~ of ch~nnels ormed.

Claims (8)

The embodiments of the invention in which an exclu-sive property or privilege is claimed are defined as follows:

1. A system for detection of an echo comprising:
a transmission array formed by M simultaneously transmitting sources El,...Em;
a reception array formed by N sources, H1,... HN;
first signal processing means for receiving the output of said reception array sources for forming P reception channels having signals V1 (.theta.,.PHI.), ... Vp (.theta.,.PHI.), where .theta.,.PHI. are spherical angles defining the receiving direction, wherein said M transmitting sources El,... Em transmit simultaneously wave signals with a duration T, said transmit-ted signals being different and separable;
a second signal processing means receiving said P
reception channel signals and outputting for each channel M at least one of decoded and filtered delayed signals, corresponding to said N trans-mitted signals, the delays of said signals corresponding to the geometry of said trans-mission sources and to the angles .theta.,.PHI. of the channel direction;
adder means for adding said delayed signals and display means for receiving the output of said adder means.

2. A system for detection as claimed in claim 1, wherein said M simultaneous transmitting sources El ... EM
transmit respectively waves of pure frequency fl, ... fM, with duration T and band b such that bT = l, separated in frequency by l/T, and wherein said second processing means includes decoding means carrying out spectral analysis.

3. A system for detection as claimed in claim 11 wherein said M simultaneously transmitting sources El,...EM
transmit waves of the same center frequency, respectively coded by said M separable codes Cl...CM of duration T and band b such that bT>M, said second processing means comprising decoding means where convoluters carry out the convolution between the signal of each reception channel such as Vk (.theta.,.PHI.) and said M
coded signals.

4. A system for detection as claimed in claim 1, wherein said M simultaneously transmitting waves have respect-ively central frequencies fl, ... fM, wherein said waves being modulated by a same code of duration T and band b such that bT
>land wherein the interval between said frequencies is b and wherein said second processing means includes decoding means carrying out a spectral analysis followed by a convolution with a copy of said same code.

5. A system for detection as claimed in claim 1, wherein said transmission signals are formed by M simultaneous signals with P frequencies fl, ... fp (P<M), each frequency being modulated by Q separable codes (PQ=M), the same Q codes being used at each of the frequencies of duration T and band b such that bT>Q, and wherein said second signal processing means includes decoding means carrying out a spectral analysis followed by a convolution with the Q copies of the codes transmitted.

6. A system for detection as claimed in claim 1, wherein said second processing means includes means for tempo-ral offset followed by a means for decoding the signals receiv-ed.

7. A detection system as claimed in claim 1, wherein said second processing means comprises for each reception channel signal, a memory containing the copy of a complex transmission signal inverted in time transmitted in the detec-tion corresponding to angles .theta. and .PHI. and wherein the signal received in channel VR (.theta.,.PHI.) is multiplied by the signal stored in said memory.

8. A system for detection as claimed in claim 1, wherein said second processing means includes a demodulator means whereby the signal supplied by the reception channels are subjected to complex demodulation with the real and imaginary components being stored in a first memory, one line corres-ponding to one reception channel, said lines being read R times at a rate Hl, placed on a carrier frequency Fl and forming the signal Vl (t) with the signals transmitted in the corresponding direction which come from a second memory having second lines being read R times at an increasing rate H2, and placed on increasing carrier frequency F2 during the R readings of said second memory in order to form a signal V2 (t) and further including convolution means for receiving said Vl (t) and V2 (t) signals and outputting, convolution signals C (t) and a demultiplexing means for receiving said convoluted signals C
(t) and applying an output to said display unit.