FR2815794A1 - Device to identify interference signals in a radio electric signal with a constant module - Google Patents

Abstract

There is a module (1), which allows to discriminate the frequency and phase of interference signals in amplitude. There is an estimation module (2) for the frequency and a calculation module (3) for the radio electric power of the interference signal.

Description

<Desc / Clms Page number 1>

modulation de fréquence ou les modulations numériques à module sensiblement constant. DEVICES FOR IDENTIFYING A BROKER SIGNAL
OF A RADIOELECTRICAL SIGNAL WITH A SUBSTANTIALLY CONSTANT MODULE
The present invention relates to devices for transmitting radio signals with a substantially constant modulus, such as the signals transmitted in

frequency modulation or digital modulations with substantially constant modulus.

 In this type of network, each constituent transmitter of the network is likely to constitute a jammer vis-à-vis a neighboring transmitter, due to the power of the required emission levels, by a phenomenon of radio interference.

 In the case of FM broadcasting, for example, the most frequent interference phenomena result from the presence of signals emitted by other transmitters of the same type. In the case of a cellular network, as shown in Figure la, this phenomenon is related to the structure of this type of network.

Thus, for a vehicle V equipped with an FM receiver, located in the overlap zone of the cells A and B, the receiver of the vehicle receives a signal emitted from the transmitter A. Such a signal can be considered as a useful signal and occupies, in principle, a frequency band of approximately 330 kHz. This frequency band, located around the carrier frequency, is usually designated per radio channel. Broadcasting Standards or Regulations, Report of the CCIR (Advisory Committee on Radio Industries) Broadcasting Service, Appendix to

<Desc / Clms Page number 2>

 Volume X, 1990, stipulate that two transmissions, from the same transmitter, must be shifted at a center frequency of at least 300 kHz from one another, a covering being introduced otherwise.

 In order to avoid overlaps in the same frequency band, the emissions of any transmitter near another transmitter, the transmitter B in figure 1a, are shifted in central frequency by 200 kHz or, if appropriate, by 100 kHz in worst case. It is conceivable, in fact, that depending on the location of the transmitters constituting the network, it is necessary to impose such small offsets, in order to meet the rules for establishing frequency allocation plans in force. An example of frequency planning is given in FIG. In the overlapping zone of two cells, the useful signal, defined as the signal emitted by the transmitter A and received by the user of the receiver, can then be scrambled by a neighboring radio channel emitted by a transmitter B, by a phenomenon adjacent channel interference.

 The aforementioned interference phenomenon occurs when the protection ratio related to the difference in emission level between the wanted signal and the interfering interfering signal is insufficient for the receiver to discriminate and correctly detect the wanted useful signal. Such a situation appears, in particular, when the receiver is located at a great distance from the transmitter of the useful signal but close to a neighboring transmitter. Under these conditions, the receiver receives a relatively weak signal from the transmitter generating the wanted signal, tuned to the

<Desc / Clms Page number 3>

 radio channel of the latter, while receiving a relatively strong signal from the transmitter transmitting on a neighboring channel.

 Indeed, although the maximum amplitude of the signal generated by the interfering emitter is not placed, in terms of frequency, in the channel of the emitter generating the useful signal, the interfering signal may however have a significant amplitude for the receiver for two reasons: either the interfering transmitter emits without strictly respecting the characteristics of the radio channel allocated to it; either the receiver is not selective enough.

 In either case, such a phenomenon causes a significant degradation of the received audio or video signal. In the current receivers, the interference signals generated by this interference phenomenon are rejected at the intermediate frequency stage.

However, such a rejection can be very insufficient to ensure the listener proper listening.

This is the case, in particular, in highly urbanized areas, where, on the one hand, the density of neighboring transmitters can be high, which consequently reduces the frequency offset of the radio channel associated with each of they and where, on the other hand, multiple reflection phenomena can, at least locally, reinforce the phenomenon of interference.

 At present, the possibilities of reducing or eliminating the aforementioned interference phenomena are reduced and poorly adapted to an effective treatment of such phenomena, the identification of the transmitter (s)

<Desc / Clms Page number 4>

 jammers, in a given site, with a view to their elimination appearing particularly difficult.

 In particular, there is currently no single-sensor type apparatus for identifying the adjacent channel interference phenomenon by digital processing. For a given frequency modulation transmitter, the identification of one or more jammers is carried out, currently, by cutoff, transmission stop neighboring transmitters and subsequent emission of these transmitters. It is thus necessary to perform a service interruption of the network.

 Such a procedure appears particularly cumbersome and uncomfortable.

 The object of the present invention is to remedy the aforementioned drawbacks of the prior art techniques, by implementing a device for identifying a jammer signal of the single-sensor type, making it possible to identify the transmitter of this interfering signal in the same manner. absence of any interruption of service of neighboring transmitters of a given transmitter constituting a substantially constant module radio signal transmission network.

 Another object of the present invention is also the implementation of a scrambling signal processing process from the identification parameters of any interfering signal delivered by the identification device, object of the invention.

 The device for identifying a jamming signal of a radio signal with a substantially constant modulus, object of the present invention, is remarkable in that it comprises a module for extracting the

<Desc / Clms Page number 5>

 interfering signal for discriminating in amplitude, in frequency and in phase this interfering signal with respect to the radio signal with a substantially constant modulus, a module for estimating the carrier frequency of this interfering signal and a module for calculating the interfering signal; radioelectric power of this interfering signal. These parameters make it possible to identify the interfering frequency, power and phase signal with respect to the substantially constant module radio signal.

 The device, object of the invention, is applied to the reception of any radio signal substantially constant module such as phase-modulated digital signals, frequency modulation signals, for example.

lesquels, outre les figures la et lb, - la figure 2 représente, à titre illustratif, un dispositif d'identification d'un signal brouilleur d'un signal radioélectrique à module sensiblement constant conforme à l'objet de la présente invention, implanté dans un circuit de réception de type FM ;

- la figure 3a représente, à titre illustratif, un mode de réalisation particulier non limitatif d'un module d'extraction de signal brouilleur constitutif du dispositif d'identification, objet de l'invention tel que représenté en figure 2 ; - la figure 3b représente, à titre illustratif, un mode de réalisation particulier non limitatif d'un module d'estimation de la fréquence porteuse du signal It will be better understood from reading the description and observing the drawings below in

which, in addition to FIGS. 1a and 1b, FIG. 2 represents, by way of illustration, a device for identifying a jamming signal of a substantially constant module radio signal in accordance with the subject of the present invention, implanted in FIG. an FM type receiving circuit;

- Figure 3a shows, by way of illustration, a particular non-limiting embodiment of a jamming signal extraction module constituting the identification device object of the invention as shown in Figure 2; FIG. 3b represents, by way of illustration, a particular nonlimiting embodiment of a module for estimating the carrier frequency of the signal

<Desc / Clms Page number 6>

brouilleur, constitutif du dispositif d'identification objet de l'invention tel que représenté en figure 2 ; la figure 3c représente, à titre illustratif, un mode de réalisation particulier non limitatif d'un module de calcul de la puissance radioélectrique du signal brouilleur, constitutif du dispositif d'identification objet de l'invention tel que représenté en figure 2 ; les figures 4a et 4b représentent, à titre illustratif, des diagrammes d'estimation de l'écart de fréquence, respectivement de l'écart-type de cet écart de fréquence entre porteuses, signal utile, signal brouilleur, pour différents émetteurs brouilleurs de canal hertzien différent, lorsque la fréquence centrale de ces émetteurs brouilleurs est estimée par la méthode du barycentre, ainsi qu'illustré en figure 3b ; les figures 4c et 4d représentent, à titre illustratif, des diagrammes d'estimation de l'écart de fréquence respectivement de l'écart-type de cet écart de fréquence entre porteuses, signal utile, signal brouilleur, pour différents émetteurs brouilleurs de canal hertzien différent, lorsque la fréquence centrale de ces émetteurs brouilleurs est estimée par la méthode du maximum, ainsi qu'illustré en figure 3b ; les figures 4e et 4f représentent, à titre illustratif, un diagramme d'estimation de la puissance du signal utile, respectivement d'un brouilleur pour différents émetteurs brouilleurs de canal hertzien différent.

jammer constituting the identification device object of the invention as shown in Figure 2; FIG. 3c represents, by way of illustration, a particular nonlimiting embodiment of a module for calculating the radioelectric power of the interfering signal constituting the identification device forming the subject of the invention as represented in FIG. 2; FIGS. 4a and 4b represent, by way of illustration, diagrams for estimating the frequency deviation or the standard deviation of this difference in frequency between carriers, wanted signal, interfering signal, for different interfering channel emitters a different radio frequency, when the center frequency of these interfering transmitters is estimated by the centroid method, as illustrated in Figure 3b; FIGS. 4c and 4d represent, by way of illustration, diagrams for estimating the frequency deviation or the standard deviation of this frequency difference between carriers, wanted signal, interfering signal, for different interfering radio-relay transmitters different, when the center frequency of these interfering transmitters is estimated by the maximum method, as illustrated in Figure 3b; Figures 4e and 4f represent, by way of illustration, a diagram for estimating the power of the wanted signal, respectively a jammer for different interfering transmitters of different radio channels.

 A more detailed description of a device for identifying a jamming signal of a radio signal with a substantially constant modulus, object of

<Desc / Clms Page number 7>

 the present invention will now be given with reference to FIG.

 In general, it is indicated that the device of the invention performs an entirely digital processing of a signal received by a receiving antenna. This processing takes place after the high frequency portion HF, the frequency transposition around the intermediate frequency HF / FI, the digital analog conversion CAN and the transformation of the real digital signal into a complex signal by the application, as represented in FIG. , of a Hilbert transform.

 The signal received at the input of the device for identifying a jamming signal, object of the present invention, is therefore a complex digital signal and in baseband, this signal thus comprising a real part and an imaginary part. For this reason, in FIG. 2, the signal delivered by the Hilbert transform module is symbolized in the device that is the subject of the present invention by a double arrow representing the aforementioned real and imaginary parts.

 As represented in FIG. 2, the device according to the invention comprising the reference 0 comprises a scrambling signal extraction module receiving the received signal and making it possible to discriminate in amplitude, frequency and phase the interfering signal of the radio signal. substantially constant module, that is to say the signal delivered by the module of the Hilbert transform. The extraction module 1 is thus deemed to deliver a scrambling signal denoted Br (n) representative of this interfering signal.

<Desc / Clms Page number 8>

 In addition, the identification device 0 comprises a module 2 for estimating the carrier frequency of the interfering signal, this estimation module receiving the discriminated interfering signal Br (n) and delivering a central frequency value of the interfering signal, this frequency value being denoted by fBr (n).

 Finally, a module 3 for calculating the radioelectric power of the interfering signal is provided, which makes it possible, from the discriminated interfering signal Br (n) delivered by the extraction module 1, to calculate the radioelectric power of this interfering signal to deliver a signal. rated power value PBr (n).

 It is thus understood that the parameters of the central frequency fBr (n) and the power PBr (n) of the discriminated interference signal thus make it possible to identify the interfering signal in frequency, in power and in phase with respect to the radioelectric signal with module substantially constant affected by the latter.

 Various particular non-limiting preferred embodiments of the extraction modules 1, the estimation 2 of the carrier frequency of the interfering signal and the calculation 3 of the radioelectric power of this same interfering signal will now be described with reference to FIGS. 3a to 3c. .

 For the preferential description of the aforementioned modules, it is indicated that the following notations are retained.

 For a received electrical signal with substantially constant modulus X (n) constituted by complex samples representative of a useful signal received Su (n), we have the relation X (n) = Su (n) = Rr (n).

<Desc / Clms Page number 9>

 In general, as shown in FIG. 3a, the scrambler signal extraction module 1 comprises a recursive predictive loop of the predictable interfering signals.

 As represented in FIG. 3a, the aforementioned recursive predictive loop comprises at least a first subtractor module 11 receiving the received radio signal with a substantially constant modulus X (n) and an estimate Yb (n) of the interfering signal signal Br (n) and delivering a discriminating received useful signal Xa (n). The subtractor module 11 thus performs the subtraction between X (n) and Br (n) or an estimate Yb (n) of this interfering signal.

 In addition, a nonlinear estimation module 12 of the received signal receives a signal Y (n) representative of the discriminated received useful signal and delivers an estimated received useful signal Z (n).

 A second subtractor module 13 is provided, which receives the estimated received useful signal Z (n) and the signal Y (n) representative of the discriminated received useful signal, this second subtracter module 13 delivering an error signal denoted E (n) = Y (n) -Z (n).

 A third subtracter module 14 is provided, which receives the estimated received useful signal Z (n) and a signal substantially representative of the substantially constant module radio signal to deliver a discriminated interfering signal S (n).

 An adaptive predictive filter 15, bearing the reference B in FIG. 3a, receives the discriminated interfering signal S (n) and delivers the estimated interference signal Yb (n) to the first subtractor module 11. The adaptive predictive filter 15 is, on the other hand, controlled by the

<Desc / Clms Page number 10>

 error signal E (n) to thereby adapt the recursive predictive loop of predictable interfering signals previously mentioned in the description.

 In addition, as shown in Figure 3a, a first 12a and a second 14a automatic gain control module are provided, these modules being driven by the error signal E (n). The first automatic gain control module 12a receives the discriminated received useful signal Xa (n) and the second automatic gain control module 14a receives the substantially constant modulus radio signal X (n). The first automatic gain control module 12a thus delivers the signal X (n) representative of the discriminated received useful signal, while the second automatic gain control module 14a delivers the signal representative of the received radio signal with substantially constant modulus to the third module. subtractor 14.

 The operation of the jammer signal extraction module Yb (n) is as follows.

Feedback Equalizer) dans laquelle la boucle prédictive récursive prédictive mise en oeuvre dans le cas du module d'extraction, objet de la présente invention, est obtenue en remplaçant la décision du processus d'égalisation PDFE par une non linéarité de Bussgang. The recursive predictive loop of predictable interfering signals can be compared with that of the PDFE (Predictive Decision) equalization algorithm

Feedback Equalizer) in which the predictive recursive predictive loop implemented in the case of the extraction module, object of the present invention, is obtained by replacing the decision of the PDFE equalization process with a non-linearity of Bussgang.

 The amplitude adjustment is performed by a recursive estimate of the useful signal module. This adaptation is slow enough not to degrade the performance of the predictor filter 15.

<Desc / Clms Page number 11>

 For a description of the scrambler cancellation process corresponding to a PDFE-type equalization process, reference may be made to the patent application filed in France on February 19, 1998 under the number 98 02021 in the name of TELEDIFFUSION OF FRANCE.

Relation 1

fxr (n) = X (n) = Su (n) + Br (n) txa (n) tsu (n)

Compte tenu de la relation précédente, il vient : Yb (n) sensiblement égal à Br (n). In the context of the implementation of the extraction module 1 shown in FIG. 3a, in accordance with the subject of the present invention, the equations making it possible to discriminate the interfering signal are given by the following relationships:

Relationship 1

fxr (n) = X (n) = Su (n) + Br (n) txa (n) tsu (n)

Given the previous relation, it comes: Yb (n) substantially equal to Br (n).

 Thus, at the output of the predictive filter 15, only the terms of the scrambler at each sampling instant n are obtained.

 With respect to the scrambler cancellation device described in the aforementioned French patent application 98 02021, the module 1 for extracting the interfering signal, object of the present invention, as represented in FIG. 3a, substantially corresponds to the device described in the above-mentioned French patent application in which the transverse filter R (z) shown in FIG. 5 and finally all the transverse part that is useful for the undesirable interfering signals is suppressed.

 Consequently, the unwanted signal extraction module 1 used for the implementation of the identification device which is the subject of the present invention makes it possible

<Desc / Clms Page number 12>

 to estimate the predictable interfering signals and to recover the estimate of the interfering signal Yb (n) which, of course, makes it possible to isolate or extract the aforementioned interfering signal.

 For a more detailed description of the module 1 for extracting the interfering signal represented in FIG. 3a and making it possible to implement the identification device that is the subject of the present invention, it will be possible to refer to French patent application No. 98 02021 previously cited. This extraction module 1 of the interfering signal thus appears as a device of the PDFE equalizer type in which the transverse part useful for the unwanted interfering signals, and therefore not necessary in the context of the implementation of the device object of the invention , has been suppressed in accordance with the physical principle of superposition of the equilibrium states relating to the processing of the signals that can not be predicted respectively from the predictable signals.

N 1. Calcul de : Xa (n) =X (n) - Lbj (n) s (n-j) j=l With regard to the actual operating mode of the extraction module 1 of the interfering signal, the signals implemented, ie signal Xa (n) delivered by the first subtractor module 11, the error signal E (n) ) and the update protocol of the predictive filter 15, verify the following relations: Relation 2

N 1. Calculation of: Xa (n) = X (n) - Lbj (n) s (nj) j = l

<Desc / Clms Page number 13>

2. Calcul de l'erreur : E (n) = Y (n) l--LY (n) J ,.....,... B (n+1) =B (n) +8b (n) E (n) S* (n) 3. Mise a-jour des flltres : a (n + 1) = a (n)-8aE (n) xa * (n)

Dans les relations précédentes, on indique que : bj (n) désigne la valeur des coefficients du filtre prédictif, a (n) désigne la valeur du coefficient d'atténuation délivrée par les premiers 12a et deuxième 14a modules de contrôle automatique de gain pour chaque échantillon de rang n, Öa et ab désignent des coefficients de pondération.

2. Computation of the error: E (n) = Y (n) l - LY (n) J, ....., ... B (n + 1) = B (n) + 8b (n ) E (n) S * (n) 3. Updating the fllters: a (n + 1) = a (n) -8aE (n) xa * (n)

In the above relations, it is indicated that: bj (n) denotes the value of the coefficients of the predictive filter, a (n) denotes the value of the attenuation coefficient delivered by the first 12a and second 14a automatic gain control modules for each sample of rank n, Öa and ab denote weighting coefficients.

 This gives a structure whose complexity is substantially identical to that of the conventional CMA process.

 The prediction process can also be normalized by dividing the adaptation steps by the signal strength.

 As regards the module 2 for estimating the carrier frequency of the interfering signal, the latter, in a nonlimiting advantageous embodiment such as represented in FIG. 3b, may comprise a module 20 for calculating the discrete Fourier transform. of the estimated interfering signal Yb (n), this calculation module delivering the spectral components of the estimated interfering signal noted Yb (k), k denoting the rank of the spectral component considered. The calculation module of the discrete Fourier transform 20 can be implemented from a fast Fourier transform calculation module for example.

<Desc / Clms Page number 14>

According to a first mode of implementation, the module 20 for calculating the Fourier transform can be followed by a module 21 for calculating the barycenter of the spectral components of the interfering signal.

 To calculate the center of gravity, the center of gravity of the samples obtained in the spectral domain is calculated for the components of rank k on a frequency window of the half-size of the Fourier transform, fast Fourier transform FFT used.

où ai=Yb (k) représente la valeur d'un échantillon. Under these conditions, the central frequency of the jammer checks the relation (3): Relation 3

where ai = Yb (k) represents the value of a sample.

 Similarly, as shown in FIG. 3b, the module 2 for estimating the carrier frequency of the interfering signal may furthermore comprise a module 22 for selecting among the spectral components of the estimated interfering signal, that is to say among the spectral components Yb (k) of the maximum amplitude spectral component. Under these conditions, the estimated carrier frequency of the interfering signal is equal to the frequency value of the maximum amplitude spectral component.

 In FIG. 3b, this operation is noted according to relation (4):

<Desc / Clms Page number 15>

fBr (n) =suplyb (k) l

Globalement, on indique que la méthode du barycentre pour estimer la fréquence centrale du signal brouilleur donne satisfaction. Toutefois, dans certains cas, la méthode du maximum est susceptible de donner de meilleurs résultats et l'une des deux méthodes peut être utilisée à partir des modules 21 et 22 représentés à la figure 3b précédemment mentionnée. Relationship 4

fBr (n) = suplyb (k) l

Overall, it is indicated that the centroid method for estimating the interfering signal center frequency is satisfactory. However, in some cases, the maximum method is likely to give better results and one of the two methods can be used from the modules 21 and 22 shown in Figure 3b previously mentioned.

 Finally, the module 3 for calculating the radioelectric power of the interfering signal advantageously comprises, as represented in FIG. 3c, a periodic calculation module, on a determined number N of samples of the estimated unwanted signal Yb (n), of a weighted sum by this number of samples of the square of the amplitude of each estimated unwanted signal sample Yb (n).

Under these conditions, the radio frequency power of the interfering signal verifies the relation (5): Relation 5

The device, object of the invention as shown in Figures 2 and 3a to 3c, can be used for receiving a frequency modulation broadcast signal, for example. The FM signal, whose carrier frequency is around 100 MHz, is received at the receiving antenna. It is then transposed to

<Desc / Clms Page number 16>

 an IF intermediate frequency of 10.7 MHz then converted into a digital signal. This digital signal is transformed into a complex signal by a digital Hilbert transform and transposed into basebands.

 In order to evaluate the efficiency of the device for identifying an interfering signal, which is the subject of the invention, different tests and simulations have been carried out for different radio channels and with two estimation methods developed, the standard deviation of the estimates : estimation of frequency deviation and interference power estimation.

 The simulations carried out were carried out assuming that the useful signal is more powerful than the interfering signal. Indeed, in the opposite case, the estimate of the interfering signal is performed more easily.

 These tests and simulations were made from a classical music signal "Opera Carmen", a speech signal, and music of varieties of the singer Michel JONASZ, each of them being scrambled by another variety extract of singer Suzanne VEGA.

 The powers of the wanted signals and the interfering signal were normalized before demodulation and equal to 1 and 0.3 respectively. The useful signal to interfering signal C / I ratio was about 5 dB. The signal-to-noise ratio before entry into the jammer identification device object of the invention was 40 dB.

 The signal filtered by a channel portion simulated the radio-mobile channel using the Jakes method. Jakes' method proposes a process that involves

<Desc / Clms Page number 17>

 model the channel coefficients by a sum of sinusoids. To obtain each path and for these to be decorrelated, each coefficient corresponds to a temporal translation of the same initial signal.

The radio channels used in the tests implemented were: - channel 1: a direct path in the absence of noise, - channel 2: a direct path + a noise at 40 dB, - channel 3: two fixed transmitters + a delay of
15 microseconds + 40 dB noise, - channel 4: a fixed transmitter + a Rice variable transmitter + a 40 dB noise, - channel 5: a Rice variable transmitter + a Rayleigh variable transmitter + a delay of
15 microseconds + 40 dB noise.

The table below gives the values of the various parameters used to perform the simulation tests mentioned above.

<Tb>
<Tb>

? <SEP> channel <SEP> Transmitters <SEP> Amplitude <SEP> Delays <SEP> Transmitters <SEP> Amplitude <SEP> Delays <SEP> Noise
<tb> fixed <SEP> (in <SEP> s) <SEP> variables <SEP> (in <SEP> us) <SEP> (in <SEP> Db)
<tb> 1 <SEP> 0 <SEP> 0
<tb> without <SEP> channel
<tb> 2 <SEP> 1 <SEP> 1 <SEP> 0 <SEP> 0 <SEP> 40
<tb> 1 <SEP> 3 <SEP> 2 <SEP> 1 <SEP> 0.7 <SEP> 0 <SEP> 15 <SEP> 0 <SEP> .40
<tb> 1 <SEP> 4 <SEP> 1 <SEP> 1 <SEP> 1) <SEP> 0 <SEP>! <SEP> 2) <SEP> 0 <SEP> 0. <SEP> 2 <SEP> 1 <SEP> 15 <SEP> 1 <SEP> 40 <SEP> 1
<tb> 5 <SEP> 2 <SEP> 1 <SEP> 0 <SEP> 0 <SEP> 2 <SEP> 0. <SEP> 2 <SEP> 0 <SEP> 15 <SEP> 40
<Tb>

Measurements were made with zero receiver speed.

<Desc / Clms Page number 18>

 In the radio channels 2, 3, 4 and 5, added a white Gaussian additive noise centered and power such that the signal to noise ratio C / B is equal to 40 dB.

 Such a noise level corresponds to the threshold of the audible, the rate of distortion then being 9.3%.

 FIG. 4a shows the estimate of the difference in frequency between the carriers by the barycentre method with respect to the frequency difference AFcb of the carrier of the interfering signal for the radio channels 1 to 5 previously mentioned in the description.

 The centroid method yields relatively accurate estimates.

 Figure 4b represents the standard deviation of the estimation of the frequency difference between the carriers by the barycenter method. This relative accuracy is obtained regardless of the radio channel up to about 150 kHz. Beyond this, the interfering signal is cut off by the station filter, which, of course, distorts the estimates.

de fréquence du signal brouilleur AFcb, respectivement l'écart-type de l'estimation de cet écart de fréquence en fonction du même écart de fréquence par la méthode du maximum. Cette méthode donne des résultats plus précis sur toute la gamme de fréquences mais uniquement pour les canaux hertziens peu sévères, tels que le canal 1 et le canal 2 précédemment mentionnés dans la description. Pour le canal 4, en particulier, les estimations sont erronées, FIGS. 4c and 4d show the estimate of the difference in frequency between the carriers evaluated by the maximum method as a function of the difference

frequency of the interfering signal AFcb, respectively the standard deviation of the estimate of this frequency difference as a function of the same frequency difference by the method of the maximum. This method gives more accurate results over the entire frequency range but only for the mild radio channels, such as channel 1 and channel 2 previously mentioned in the description. For channel 4, in particular, the estimates are wrong,

<Desc / Clms Page number 19>

 which tends to prove that the maximum method is less robust than the centroid method.

 Thus, the extraction module 1 previously described in the description of the interfering signal implemented in the device for identifying a jamming signal which is the subject of the invention appears as a good estimator of the normalized power of the wanted signal with a small signal. restriction for channel 4 previously described. The device, object of the invention, can be used advantageously in a multi-jitter configuration. Under these conditions, the extraction module 1 is able to distinguish a useful signal in the middle of several jammers. In addition, this extraction module 1 appears particularly suitable for increasing the number of loops, to identify several jammers for example.

 The operating mode of the scrambling signal extraction module 1, which is the subject of the present invention, makes it possible to identify any interfering signal FM whatever its frequency offset above 25 kHz with respect to the carrier frequency of the signal. useful.

Claims (6)

 means for calculating the radio frequency power of said interfering signal, which makes it possible to identify said interfering signal in terms of frequency, power and phase with respect to the substantially constant module radio signal.

 1. A device for identifying a jamming signal of a substantially constant modulus radio signal, characterized in that it comprises: means for extracting the interfering signal, for discriminating in amplitude, frequency and phase said signal interfering with said substantially constant module radio signal; means for estimating the carrier frequency of said interfering signal;  2. Device according to claim 1, characterized in that for a substantially constant modulus received radio signal X (n) consisting of complex samples representative of a received useful signal Su (n) and said interfering signal Br (n) , X (n) = Su (n) + Br (n), said interfering signal extraction means comprise a recursive predictive loop of predictable interfering signals.  3. Device according to claim 2, characterized in that said recursive predictive loop comprises at least: a first subtractor module receiving said received radio signal substantially constant module X (n) and an estimate Yb (n) of said interfering signal Br (n) and delivering a discriminated received useful signal Xa (n); <Desc / Clms Page number 21> Y (n) representative of said discriminated received useful signal, respectively said signal representative of said received radio signal with substantially constant modulus.  nonlinear estimation means of the received useful signal receiving a signal Y (n) representative of said discriminated received useful signal and delivering an estimated received useful signal Z (n); a second subtractor module receiving said estimated received signal Z (n) and said signal Y (n) representative of said discriminated received useful signal and delivering an error signal E (n) = Y (n) -Z (n); a third subtractor module receiving said estimated received useful signal Z (n) and a signal representative of said substantially constant module radio signal and delivering a discriminated interfering signal S (n); an adaptive predictor filter, driven by said error signal E (n), receiving said discriminated interfering signal S (n) and delivering said estimated interfering signal Yb (n) to said first subtracter module; first and second automatic gain control means, driven by said error signal receiving said discriminated received useful signal Xa (n), respectively said substantially constant modulus radio signal X (n), and delivering said signal  4. Device according to one of claims 1 to 3, characterized in that said means for estimating the carrier frequency of said interfering signal comprise at least: means for calculating the Fourier transform of said estimated interfering signal Yb (n) , delivering spectral components of said estimated interfering signal; <Desc / Clms Page number 22>  means for calculating the barycentre of said spectral components, the estimated carrier frequency of said interfering signal being taken equal to the value of the barycenter of said spectral components.  5. Device according to one of claims 1 to 4, characterized in that said means for estimating the carrier frequency of said interfering signal comprise at least: means for calculating the Fourier transform of said estimated interfering signal Yb (n) , delivering the spectral components of said estimated interfering signal; means for selecting, from among said spectral components of this estimated interference signal, the maximum amplitude spectral component, the estimated carrier frequency of said interfering signal being taken equal to the frequency value of said maximum amplitude spectral component.  6. Device according to one of claims 1 to 5, characterized in that said means for calculating the radioelectric power of said interfering signal comprises periodic calculation means, on a determined number N of samples of said estimated unwanted signal Yb (n ), a sum weighted by this number of samples, of the square of the amplitude of each sample of said estimated interfering signal Yb (n).