EP0903960A1 - Antenna formed by a plurality of acoustic detectors - Google Patents

Abstract

A mathematical formula is followed for the constraints which involves a pure delay and reduction of the incoherent noise level, providing improved near field performance.

Description

The present invention relates to an acoustic antenna formed from a plurality of discrete acoustic transducers, in particular an acoustic reception antenna, that is to say formed of a plurality of acoustic sensors or microphones. Being given the principle of reciprocity, the invention also applies to an antenna emission acoustics.

The main purpose of an acoustic receiving antenna is to reduce all defects in reception while retaining useful information, i.e. information issued by the speaker or by the useful source.

Below, to better appreciate the difficulties that the invention aims to overcome, we will develop a classic theoretical study of antenna arrays acoustic, taking the case of an antenna with arbitrary geometry, composed of acoustic sensors that have arbitrary directivity patterns.

The acoustic signals received on the antenna sensors are damaged by: (1) other issuers; (2) multi-path propagation; (3) in some cases, a echo; (4) electronic noise from sensors and amplifiers; and (5) possibly, the quantization noise for digital processing.

We suppose a linear and additive model, that is to say that the degradations not are not taken into account. In the following, the disturbances (1) to (3) will be say "spatially coherent" or simply "coherent" while the disturbances (4) and (5) are said to be "inconsistent".

The performance of an antenna against a coherent disturbance is given by its directivity diagram. The speaker is assumed to be located in the near field, this which means that instead of being interested in a direction, we are rather interested in a point in the space. We assume that the coherent disturbance sources are in the field distant.

We have retained a formula which expresses the improvement in the signal to coherent disturbances, under the assumption of a diffuse field in comparison with a omnidirectional sensor placed at the location of the sensor closest to the antenna. The reflections are treated as image sources. So you just need to know the law of propagation in free field as well as the directivity diagram of each sensor.

où

x_m

signal du capteur m, aussi nommé observation

t

temps

u_p,m

directivité du capteur m dans la direction de la source p

s_p

signal émis par la source p

d_p,m

distance source p - capteur m

c

vitesse de propagation

b_m (t)

bruit incohérent (bruit électrique et de quantification) sur le capteur m

A typical model for propagation is:

or

x _m

sensor signal m, also called observation

t

time

u _{p, m}

directivity of the sensor m in the direction of the source p

s _p

signal from source p

d _{p, m}

source distance p - sensor m

vs

propagation speed

b _m ( t )

inconsistent noise (electrical and quantization noise) on the m sensor

où
   X, S, B observation, signal émis et bruit dans le domaine fréquentiel f    fréquenceTo simplify the calculation, we pass into the frequency domain:

or
X, S, B observation, signal emitted and noise in the frequency domain f frequency

Antenna processing can be seen as a dot product in the frequency domain. The signal at the end of the processing is expressed in the form:

Suppose that the useful source is the source p = 1. A conventional antenna processing consists in putting the signal back in phase, possibly weighting the sensors to establish a compromise between the opening of the main lobe and the level of the secondary lobes, and in calculating this sum. We can express it by a set of coefficients: w m (( f ) = g m (( f ) e j 2πf vs d 1, m with g _m ( f ) real and positive

At the exit, we therefore have:

   et on obtient, avec λ = c / f:
   le gain complexe du signal utile :

   le gain complexe du signal perturbateur cohérent :

   le facteur de directivité :

The three terms of the above sum correspond respectively to the useful signal, to the coherent disturbances and to the incoherent noise. This equation can be used for arbitrary linear processing if complex values for g _m ( f ) are allowed. To obtain the directivity factor, we must vary the position of a disturbing source, ie p = 2, and calculate the average of the residual of the disturbing signal. We first introduce an amplitude factor, the last term of which is used to obtain a factor independent of the distance if it is sufficiently large:

and we obtain, with λ = c / f :
the complex gain of the useful signal:

the complex gain of the coherent disturbing signal:

the directivity factor:

et d 2(f) = α2,1 e j2π d 1,1-d 2,1 λ ,...,α2,M e j2π d 1,M -d 2,M λ T    on obtient :

   et, finalement, avec les matrices A (f) = a^H ₁ (f) a₁ (f)et

   on a : F(f)= g(f)A(f)gH (f) g(f)D(f)gH (f) With the following vector notations:

and d 2 (( f ) = α 2.1 e j 2π d 1.1 - d 2.1 λ , ..., α 2, M e j 2π d 1, M - d 2, M λ T we obtain :

and, finally, with the matrices A ( f ) = a ^H ₁ ( f ) a ₁ ( f ) and

we have : F (( f ) = g (( f ) AT (( f ) g H (( f ) g (( f ) D (( f ) g H (( f )

As already indicated, these equations are based on a propagation model which is very well adapted in free field without obstacles. To adapt the calculation to a situation in which the model does not prove to be sufficiently precise, the propagation model can be replaced by measurements. In this case, the vectors d ₂ ( f ) represent measured propagation vectors.

et F U (f) = g(f)A(f)gH (f) g(f)DU (f)g H (f) We can generalize this result by introducing a weighting U ( f, ϕ,  ) of the quadratic error of the integral along the direction:

and F U (( f ) = g (( f ) AT (( f ) g H (( f ) g (( f ) D U (( f ) g H (( f )

It is assumed that the incoherent noise is not correlated from one sensor to another and that its power is equal to σ 2 / b ( f ) for all the sensors. The reduction of inconsistent noise is written in this case:

   et on peut tracer le diagramme de directivité Ω_f,ϕ0 (ϕ) pour une fréquence donnée en faisant varier ϕ:

From this study, we can deduce the classical delay-weighting-summation treatment by focusing in the far field. For a rectilinear antenna with uniform spacing d of the sensors, the complex gain of the coherent disturbing signal G2 becomes:

and we can draw the directivity diagram Ω _{f, ϕ0} (ϕ) for a given frequency by varying ϕ:

This classic treatment has been the subject of numerous studies since 1946. We knows the method of C.L. Dolph described in the technical journal "Proceedings of the I.R.E. on Waves and Electrons ", Vol. 34, n ° 6, June 1946, pp. 335-348. In this method, we space the sensors equidistantly and we adjust their sensitivities in accordance with the coefficients of the Tchebycheff polynomials so as to obtain a response comprising a main lobe of a given level and several secondary lobes of lower levels, practically equal. Since we only use fractions of sensibilities of the sensors, the network produces a response which has a signal / noise ratio lower than it would be if we used the full sensitivity of each sensor. Else hand, if the distance between the sensors is too large or too small compared to the wavelength, antenna performance drops.

More recently, document FR-A-2 472 326 describes a method of optimization of a linear geometry of acoustic antenna, with conventional summation of the signals sensors. We can consider that it is a linear delay-sum antenna at variable spacing. This antenna only works well around a frequency in narrow band and the antenna is relatively large compared to the wavelength.

Even more recently, the document FR-A-2722637 describes a geometry antenna in which the sensors are distributed in a horizontal plane on a line concave towards a speaker. We add the signals from the sensors in phase. The antenna is decomposed into sub-antennas each characterized by a specific spacing between sensors and each assigned to a part of the frequency band. In bass frequencies, we always encounter difficulties.

Conventional treatments of this type have been studied by other researchers who have chosen different weights to modify the opening of the lobe main and the level of the secondary lobes of the directivity diagram. To note that, in these treatments, the directivity diagrams of the sensors are not used.

When the antenna is to receive broadband acoustic signals, i.e. including frequencies as low as 20 Hz, we find, with the processing conventional, two difficulties: a necessarily high number of antenna sensors and a large dimension of the antenna. Classic treatments therefore lead to expensive and cumbersome solution.

As an alternative, a so-called "superdirective" antenna processing has been proposed in which the directivity factor is optimized. On this subject, reference may be made to the book "Antenna Handbook" edited by YT Lo and SW Lee in 1993, Vol. II, chapter 11 entitled "Array Theory" and in particular on pages 11-61 to 11-79 of this chapter 11. According to the present study exposed above, the maximization of the directivity factor (relation 5) for a source in far field (the α are all equal to 1) is expressed starting from relations 4 and 5 by: g opt (( f ) = arg max g (( f ) D (( f ) g H (( f ) g (( f ) AT (( f ) g H (( f ) and, by fixing a transfer function equal to unity in the direction of the useful signal, by the constraint: g (( f ) at H 1 (( f ) = 1

By this treatment, we can reduce the distance between sensors which becomes more small compared to the wavelength. Good spatial selectivity is thus obtained with a small antenna. The disadvantages of this antenna superdirective are the weak robustness, i.e. a rapid decrease in performance if the optimization is not perfect or if we deviate from the conditions optimal use; amplification of inconsistent noise; and the drop in performance when the information does not come from the end-fire management.

Among the recent works relating to end-fire acoustic antennas, we can cite the article entitled "Practical supergain" by H. Cox et al, published in "IEEE Transactions on Acoustic Speech and Signal Processing ", Vol. ASSP-34, n ° 3, June 1986, pp. 393-398. This superdirective antenna is always optimized to aim in far field, because the module is not used. In addition, there are no constraints possible linear lines and the directivity of the sensors is still not taken into account. The weighting is only subject to a constraint on the gain compared to the white noise not correlated.

We still tried to improve performance using algorithms adaptive which allow to estimate the field and to follow its evolution. The results are satisfactory if the following three conditions are met: (1) the number of sources must be small compared to the number of sensors; (2) the background noise is more energetic than the indirect paths of the useful source; and (3) the variation of the field is not too fast. If the first condition is not met, it is difficult analyze the field because of the ambiguities. The second condition is necessary for do not confuse the disturbing signal to be minimized with the useful signal. The third condition is necessary so that the algorithm can follow with an adaptation step small enough to avoid unstable behavior.

Based on these basic treatments, all valid in the far field: treatments classical, superdirective, adaptive algorithms, we tried to develop a delay-weighting-summation lobe formation treatment by focusing in near field. Instead of equalizing delays for one direction, we equalize delays for a point in the near field. But, while the known treatments mentioned previously, are well understood, because the directivity diagram can be expressed by the Fourier transform of the weighting, few satisfactory results have been published for near field focusing.

In the article "Near-Field Beamforming for Microphone Arrays" by J.G. Ryan and R.A. Goubran, published in "Procceedings of IEEE ICASSP", 1997, pp. 363-366, we take into account the term 1 / R for the attenuation and therefore we use the module signals. We still use a rectilinear geometry of conventional antennas evenly spaced. But we do not integrate the directivity diagram of the sensors. In addition, as we will see below, we optimize a function which depends on signals to be processed and no additional linear constraints are integrated.

Indeed, the treatments mentioned so far do not solve certain difficulties because, on the one hand, the sound signals to be processed belong to a frequency spectrum to broadband, occupying several octaves, for example from 100 to 8000 Hz and, on the other hand, there are near-field sound sources for which is not verified the propagation of sound waves by plane waves. In particular, a small conventional antenna cannot be selective at low frequencies.

An object of the present invention is to provide an antenna processing which improves existing conventional treatment, starting with treatment of superdirective genre in which the module is processed so as not to distort useful signal from a near-field acoustic source which respects a number of constraints.

Another object of the invention consists in providing an antenna composed of a plurality of acoustic sensors whose output signals are processed, the signal processing output being higher in quality than the output signal from a state antenna of the technique when the useful acoustic source is in the near field.

Another object of the invention consists in providing, an antenna whose processing provides better selectivity at low frequencies.

• un facteur de directivité élevé,
• un signal utile peu distordu, et
• une réduction élevée du bruit incohérent.

Another object of the invention consists in providing an antenna having:

• a high directivity factor,
• an undistorted useful signal, and
• a high reduction in inconsistent noise.

According to a characteristic of the present invention, there is provided an antenna formed by a plurality of acoustic sensors whose sensor output signals are subjected to a superdirective type of processing, with a constraint with regard to the module and a non-constraint. linear which fixes the reduction of the inconsistent noise, the theoretical formulation of these constraints being the following: g (( f ) at H 1 (( f ) = e -j2πfτ and g (( f ) g H (( f ) < 1 R min the first constraint signifying that the total transfer function is a pure delay τ, and the second constraint signifying that a limit is set for the reduction of the inconsistent noise.

According to another characteristic, the processing of said antenna is still subject to another constraint signifying, for example, the presence of one or more zeros of the directivity diagram in one or more given directions, that is to say: VS (( f ) g H (( f ) = p (( f )
where C ( f ) is a matrix of propagation vectors,
and
p ( f ) is a complex gaiin vector for each propagation vector.

According to another characteristic, said processing is carried out by an operator mathematical in a flowchart called superdirective-module-phase or SDMP whose the input data is the antenna geometry and model data propagation, weighting data and stress data mentioned above, and whose output data are, in the frequency domain, the coefficients of a plurality of digital filters as numerous as the sensors acoustic.

According to another characteristic, there is provided an antenna formed of a plurality acoustic sensors, a first part placed in front of a useful source close consists of sensors aligned in a first row and a second part placed behind the first row in relation to the near useful source consists of sensors aligned in at least a second row.

According to another characteristic, the common direction of the rows of sensors in the first and second parts are transverse to the mean direction of useful acoustic waves.

According to another characteristic, the common direction of the rows of sensors in the first and second parts are slightly oblique to the direction mean useful acoustic waves.

According to another characteristic, the sensors of the first part are distributed symmetrically in a logarithmic fashion around the median sensor.

According to another characteristic, the sensors of the first part are selectively assigned to several sub-antennas, each sub-antenna being associated with a determined band of frequencies and the sensors selectively assigned to this sub-antenna delivering output signals which are processed by conventional processing, the frequency bands being contiguous and the whole of them not falling below practically 1 kHz, each processing consisting of a specific filtering and the the output signals of each specific filter being summed.

According to another characteristic, in the antenna, each output signal of a sensor is filtered by a filter that performs, at the same time, the SDMP algorithm for bass frequencies, cutting into frequency bands according to the antenna method logarithmic, and classical channel formation for frequencies not processed by the SDMP algorithm.

According to another characteristic, a propagation model is used.

According to another characteristic, a measure of the vectors of spread.

la Fig. 1 est un schéma illustrant le traitement des signaux de sortie des capteurs acoustiques d'une antenne quelconque de l'invention,

la Fig. 2 est une vue schématique d'un premier exemple d'antenne suivant l'invention,

les Figs. 3 et 4 représentent respectivement deux diagrammes de modules et deux diagrammes de différence de phase concernant les filtres utilises dans l'antenne de la Fig. 2,

la Fig. 5 est un schéma synoptique d'un circuit de traitement des signaux de sortie des capteurs de l'antenne de la Fig. 2,

la Fig. 6 représente schématiquement trois courbes de réponse en fonction de la fréquence qui sont obtenues suivant trois hypothèses différentes,

la Fig. 7 est une vue schématique d'un deuxième exemple de réalisation d'une antenne en U suivant l'invention,

la Fig. 8 est le schéma synoptique d'un circuit de traitement des signaux de sortie des capteurs de l'antenne de la Fig. 7,

la Fig. 9 est une vue schématique d'un troisième exemple de réalisation d'une antenne en Pi suivant l'invention, et

la Fig. 10 est une vue schématique d'un quatrième exemple de réalisation d'une antenne en T suivant l'invention.

The characteristics of the present invention mentioned above, as well as others, will appear more clearly on reading the description below of exemplary embodiments, said description being made in relation to the accompanying drawings, among which:

Fig. 1 is a diagram illustrating the processing of the output signals from the acoustic sensors of any antenna of the invention,

Fig. 2 is a schematic view of a first example of an antenna according to the invention,

Figs. 3 and 4 respectively represent two module diagrams and two phase difference diagrams concerning the filters used in the antenna of FIG. 2,

Fig. 5 is a block diagram of a circuit for processing the output signals of the sensors of the antenna of FIG. 2,

Fig. 6 schematically represents three response curves as a function of frequency which are obtained according to three different hypotheses,

Fig. 7 is a schematic view of a second embodiment of a U-shaped antenna according to the invention,

Fig. 8 is the block diagram of a circuit for processing the output signals of the sensors of the antenna of FIG. 7,

Fig. 9 is a schematic view of a third embodiment of a Pi antenna according to the invention, and

Fig. 10 is a schematic view of a fourth embodiment of a T-antenna according to the invention.

Fig. 1 symbolically shows the SDMP 10 organization chart which receives input data of a set 11 containing the digital data relating to the topographic location of the antenna sensors as well as the useful source, a set 12 containing the data relating to the linear constraints, of a set 13 containing the data relating to the spatial weighting, of a set 14 containing the data relating to the constraints on the chosen reduction of the inconsistent noise, and a set 15 containing the data relating to the definitions of the sub-antennas. The flowchart 10 delivers output data to a set 16, the data of output being relative to a set of coefficients of M digital filters in the domain frequency, M being equal to the number of sensors of the antenna.

A statement of the SDMP flowchart of the invention which embodies the operator above mentioned math is shown in the appendix at the end of the presence description. This flowchart is described in MATLAB language, well known to humans of career.

With the set of M filters in the frequency domain, we can proceed either filtering in the frequency domain with multiplication, or transform by a classical filter design algorithm, for example the type algorithm "generalized least squares", to obtain a set of filters in the time domain, then to filter in the time domain with convolution.

In Fig. 2, the antenna is formed by two acoustic sensors or microphones 21 and 22 placed one behind the other in relation to a useful speaker or acoustic source 23. The sensors 21 and 22 and the useful source 23 are aligned. The distance d between sensors is, for example, 30 cm and is equal to the distance from sensor 21 to the source 23. This very simple antenna thus symbolizes sound pickup in the near field. Of more, always for the sake of simplicity, it is assumed that the two sensors have a omnidirectional directivity diagram.

The outputs of sensors 21 and 22 are respectively connected to the inputs of low-pass filters 24 and 25 whose outputs are connected to the inputs of a summator 26 which outputs the antenna output signal at 27.

With a conventional treatment - "equalization of the delay due to propagation, then summation" - at very low frequencies, the coherent disturbances coming from all directions are summed in phase, which quadruples the power, either with formula (2) ci -above: | G 2 | 2 = (1 + 1) 2 = 4

The useful signal is also added in phase, but the amplitude of the signal on the sensor 2 is two times smaller than on the sensor 1, which leads to an amplification of the power of the useful signal equal to: | G 1 | 2 = (1 + 0.5) 2 = 2.25 and a directivity factor - formula (3) above - equal to:

If we perform a subtraction, instead of a summation as in classical processing, we have: | G 2 | 2 = (1-1) 2 = 0 a useful signal: | G 1 | 2 = (1 - 0.5) 2 = 2.25

Thus, the directivity factor tends to infinity if the frequency tends to zero. On the other hand, the processing is less robust, because the useful signal is weak at the output. The amplification of the signal amplifies everything that is not identical on the two sensors 1 and 2, that is to say the incoherent noise which is added in power: 1 2 +1 2 = 2 which means an amplification of the inconsistent noise compared to the useful signal equal to:

This amplification remains small compared to the infinite directivity factor. he appears that the treatment of the invention makes it possible to find a compromise between the directivity factor and inconsistent noise amplification.

• avec l'hypothèse (a), il n'y a pas de contrainte pour l'amplification du bruit incohérent,
• avec l'hypothèse (b), on accepte une amplification du bruit incohérent compris entre 0 et 5 dB, et
• avec l'hypothèse (c), on prend une réduction du bruit incohérent égale à la solution classique, c'est-à-dire
  
  .

Three treatments according to the invention were examined in different hypothetical cases:

• with hypothesis (a), there is no constraint for the amplification of the inconsistent noise,
• with hypothesis (b), we accept an amplification of the inconsistent noise between 0 and 5 dB, and
• with hypothesis (c), we take a reduction in the incoherent noise equal to the classical solution, i.e.
  
  .

In hypothesis (a), low-pass filters 24 and 25 are used whose diagrams modules as a function of frequency are respectively indicated in FIG. 3. We can see that for f = 0, the amplitudes of the two modules are equal, which is justified by the equalities above. Above 400 Hz, the amplitudes decrease significantly by -4 dB to reach -12 dB for filter 24 and -18 dB for filter 25.

Still in hypothesis (a), to highlight the components of the signal useful the diagrams of phase differences as a function of frequency, Fig. 4, in taking into account the delays, indicate that the responses of filters 24 and 25 are in phase opposition for f = 0, but are practically of the same value beyond 400 Hz.

The block diagram of FIG. 5 shows an embodiment of a processing - filtering, summation - at the output of sensors 21 and 22 in the field temporal. The outputs of sensors 21 and 22 are respectively connected to the inputs microphone amplifiers 28 and 29 whose outputs are respectively connected at the inputs of analog-digital converters 30 and 31 whose outputs are respectively connected to the memory inputs 32 and 33 made up of registers with shift comprising, for example, thirty-two cells each. The lateral exit of a memory cell 30, associated with sensor 24, is connected to a door input 34.1.n whose second input receives a signal of coefficient h.l.n. The lateral exit of a memory cell 31, associated with sensor 25, is connected to a door entry 34.2.n whose second input receives a signal of coefficient h.2.n. The parameters n mentioned above vary discreetly from one to thirty-two depending on the rank of the cell in the shift register. The outputs of doors 34.1.n and 34.2.n are connected to the corresponding inputs of a digital summator 26 whose output delivers at 27 the antenna signal.

In Fig. 6, the variation of the directivity factor as a function of the frequency, in hypothesis (a), is indicated by the curve al, which decreases from 25 dB to 5 dB below of 100 Hz, shows that we improve performance at low frequencies compared to those of a conventional antenna indicated by the curve Id. The curve 2a indicates the variation of the reduction.

Still in FIG. 6, in hypothesis (b) where we accept an amplification of the inconsistent noise between 0 and 5 dB, curve 1b shows that the low frequency performance up to 5 dB, i.e. where solutions classics don't work well. Curve 2b corresponds to the variation of the minimum reduction imposed.

Finally, in hypothesis (c) where we took a reduction in the inconsistent noise equal to the classic solution, curve 1c shows that we can gain between 2 dB for low frequencies and 0.6 dB for high. Line 2c identical to line 2d corresponds to the variation of the minimum reduction imposed.

We note, in these three hypotheses, that the antenna is all the less directive that the reduction in the inconsistent noise is stronger, that the algorithm of the invention gives better results than the conventional solution 1d and 2d by comparing the curves 1c and 1d, and that the directivity factor can be high for low frequencies.

We can therefore choose a compromise between reduction of inconsistent noise and factor of directivity.

In Fig. 7, there is shown schematically a useful source 100 U-shaped antenna comprising thirteen sensors 101 to 113 which in the example described are forward directional cardioid pattern sensors, i.e. the region containing source 100 relative to the antenna. The first nine sensors 101 at 109 are aligned symmetrically around the sensor 105 on a first straight line D1, the two following sensors 110 and 111 are arranged on a second straight line D2 and the two last sensors 112 and 113 on a third straight line D3. Lines D1, D2 and D3 are parallel and perpendicular to a straight line D4 passing through the sensor 105 and on which the useful source 100 is installed. For example, the distance from the source 100 to the right D1 is 60 cm and the lines D2 and D3 are respectively placed behind the right D1 at 15 and 30 cm. The sensors 110 and 112 are aligned behind the sensor 101 and sensors 111 and 113 are aligned behind sensor 109 so that train the legs of the U.

On the right D1, the intervals between the sensors 105, 104, 103, 102 and 101 vary increasing logarithmically and symmetrically to intervals between sensors 105, 106, 107, 108 and 109.

Between 105 and 104, the interval is 2.5 cm; between 104 and 103, it is 2.5 cm; between 103 and 102, 5 cm; and between 102 and 101, 10 cm The sensor 110 is placed 15 cm behind sensor 101, like 111 behind 109, and sensor 112 is placed 15 cm behind the sensor 110, like 113 behind 112.

The block diagram of FIG. 8 illustrates the frequency implementation of the filtering output signals from sensors 101 to 113 in Fig. 7. The sensor 101 supplies a amplifier A01 followed by an analog-digital converter B01 followed by a circuit C01 operating according to the Fast Fourier transformation algorithm (TFR with zero padding) connected to the serial input of a D01 filter whose output is connected to an input corresponding of an SOM adder. The parallel input of filter D01 receives the clearance coefficients calculated by the SDMP flowchart for this filter.

In Fig. 8, we have included the sensor 113 which supplies an amplifier A13 followed by an analog-digital converter B13 followed by a circuit c13, operating like circuit C01, connected to the serial input of a filter D13, the output of which is connected to a corresponding input of the SOM adder. The parallel input of filter D13 receives also a set of coefficients calculated by the SDMP organization chart.

The output of the SOM adder is connected to a circuit E operating according to a Reverse Fast Fourier Transformation algorithm (TFRI with Overlap Add) followed by a digital analog converter F which delivers the output signal from the antenna. .

In practice, the algorithm can be performed in real time using a DSP (Texas Instruments C50).

In practice, for the treatment, the antenna of FIG. 7 in four sub-antennas, including the first three, in which the sensors 101 to 109 operate from the line D1, are used to cover three octaves at high frequencies and the fourth in which all the sensors 101 to 113 intervene is used for cover low frequencies from 0 to 1 kHz.

As mentioned above, on the right D1, the sensors 101 to 109 are symmetrically distributed in a logarithmic manner, which allows in a way known per se to reduce the number of sensors, here to nine. A number of five sensors per octave band is sufficient. The sensors 103 to 107 are used, constituting the first sub-antenna, for the band 4 to 7 kHz; sensors 102, 103, 105, 107 and 108, constituting the second sub-antenna, for the band 2 to 4 kHz; and the sensors 101, 102, 105, 108 and 109, constituting the third sub-antenna, for the band 1 to 2 kHz.

In the fourth sub-antenna, the processing involves all the sensors 101 to 113 using the algorithm of the invention, that is to say taking into account the module differences and phase differences on sensors 110 to 113, of a similar to the processing mentioned above for the antenna of FIG. 2.

Thus the treatment according to the invention is useful for a wide band of frequencies, for example for speech a band going from 20 Hz to 7 kHz.

In Fig. 9, a variant of the antenna of FIG. 6 comprises, opposite a source useful 200, thirteen sensors 201 to 213 with directivity diagram in cardioid. The nine first sensors 201 to 209 are symmetrically aligned around the sensor 205 on a first straight line D1, the two sensors along 210 and 211 are arranged on a second line D2 and the last two sensors 212 and 213 on a third line D3. Lines D1 to D3 are parallel and perpendicular to a line D4 passing through the sensor 205 and the useful source 200. In the example shown, the mutual distances between the lines D1 to D3 and the source 200 are identical to those mentioned in About the antenna of Fig. 6.

On the right D1, the mutual distances between the sensors 201 to 209 are identical to those that exist between sensors 101 to 109.

Sensors 210 and 212 are aligned behind the middle of segment 201-202 and the sensors 211 and 213 aligned behind the middle of the segment 208-209. In depth, their mutual distances are the same as in FIG. 7. The offsets of the sensors 210 at 213 towards the center of the antenna earns it the antenna designation in Pi.

Pi antenna output signals are processed according to the flowchart superdirective-module-phase of the invention.

In Fig. 10, another variant of the antenna of FIG. 6 has in front of a useful source 300, thirteen sensors 301 to 313 with a cardioid directivity diagram. The first nine sensors 301 to 309 have, on the right Dl, the same arrangement as the first nine sensors of FIG. 6.

The last four sensors 310 to 313 are successively aligned according to the same line D4 in FIG. 6, behind 305 so as to form, with the sensors 301 to 309, a T-shaped antenna. The distance between the sensors 310 and 305 is 10 cm, as between sensors 311 and 310, between 312 and 311, and between 313 and 312.

T-antenna output signals are processed according to the flowchart superdirective-module-phase of the invention.

In variants, instead of giving the antennas in U, Pi or T, described above in relation to Figs. 7, 8 or 9, a straight structure, we can give them an oblique structure, that is to say that the lines Dl, D2, D3 are no longer perpendicular to the line D4, but make a certain angle with it, the position of the useful source being always aligned with the straight line D4.

In Fig. 1, there is shown a set 11 which contains the digital data relating to the topographic location of the antenna sensors as well as the useful source. This set 11 still contains data relating to the propagation and / or, as mentioned above, measures of responses impulse.

In the appendix below, there is indicated, as already mentioned, an SDMP flow chart written in MATLAB language.

Claims (11)

Antenna formed by a plurality of acoustic sensors characterized in that the sensor output signals are subjected to a superdirective type of processing, with a constraint with regard to the module and a non-linear constraint which fixes the reduction of incoherent noise, the theoretical formulation of these constraints being as follows: g (( f ) at H 1 (( f ) = e -j2πfτ
and g (( f ) g H (( f ) < 1 R min the first constraint signifying that the total transfer function is a pure delay τ, and the second constraint signifying that a limit is set for the reduction of the inconsistent noise. Antenna according to claim 1, characterized in that said processing is still subject to another constraint signifying the presence of one or more zeros of the directivity diagram in one or more given directions, that is to say: VS (( f ) g H (( f ) = p (( f ) where C (f) is a matrix of propagation vectors,
and
p (f) is a complex gain vector for each propagation vector. Antenna according to claim 1 or 2, characterized in that said processing is materialized by a mathematical operator in a so-called superdirective-module-phase flowchart or SDMP whose input data is the geometry data of antenna and propagation model, weighting data and data relating to the constraints mentioned above, and whose output data are, in the frequency domain, the coefficients of a plurality of digital filters also as many as acoustic sensors. Antenna according to claim 1 to 3, characterized in that it is formed a plurality of acoustic sensors, a first part of which is placed opposite a near useful source consists of sensors aligned in a first row and a second part placed behind the first row in relation to the near useful source is consists of sensors aligned in at least a second row. Antenna according to claim 4, characterized in that the direction common rows of sensors in, the first and second parts are transverse to the mean direction of the useful acoustic waves. Antenna according to claim 4, characterized in that the direction common rows of sensors in the first and second parts are slightly oblique to the mean direction of the useful acoustic waves. Antenna according to one of Claims 4 to 6, characterized in that the sensors of the first part are distributed symmetrically in a way logarithmic around the median sensor. Antenna according to claim 7, characterized in that the sensors of the first part are selectively assigned to several sub-antennas, each sub-antenna being associated with a determined frequency band and the sensors selectively assigned to this sub-antenna delivering output signals which are treated by a conventional treatment, the frequency bands being contiguous and their together not falling below practically 1 kHz, each processing consisting of specific filtering and the output signals of each specific filter being summoned. Antenna according to claim 8, characterized in that each signal of output of a sensor is filtered by a filter that performs, at the same time, the SDMP algorithm for low frequencies, cutting into frequency bands according to the antenna method logarithmic, and the classical channel formation for the frequencies not treated according to the SDMP algorithm. Antenna according to any one of Claims 1 to 9, characterized in that that we use a propagation model. Antenna according to any one of Claims 1 to 9, characterized in that that we use a measure of the propagation vectors

Images