EP1423727A1

EP1423727A1 - Interference suppression method and device, in reception, for a wideband radio signal

Info

Publication number: EP1423727A1
Application number: EP02783140A
Authority: EP
Inventors: Valéry Thales Intellectual Property LEBLOND; Alain Thales Intellectual Property RENARD; Franck Thales Intellectual Property LETESTU
Original assignee: Thales SA
Current assignee: Thales SA
Priority date: 2001-09-07
Filing date: 2002-09-03
Publication date: 2004-06-02
Also published as: FR2829638A1; US7453923B2; US20040234016A1; WO2003023445A1; CA2459369A1; FR2829638B1

Abstract

The invention relates to an interference suppression method and device, in reception, for a wideband radio signal. The invention involves the use of spatial filtering taking into account different interfered versions of the signal delivered by parallel processing chains which are believed to have the same transfer function, such as, for example, the individual processing chains associated with the radiant elements of an array antenna that receives GPS navigation satellite signals. The inventive method consists in performing an automatic equalisation (4d) which reduces the disparities between the processing chains (2a to 2g) regardless of whether said disparities are original or due to drifts over time. Said automatic equalisation is performed at the outputs of the processing chains (2a to 2g) before the spatial interference suppression filtering (3) and it is carried out by means of a local test signal generator (5).

Description

METHOD AND DEVICE FOR ANTI-INTERFERENCE, IN RECEPTION, OF A BROADBAND RADIOELECTRIC SIGNAL

The present invention relates to interference suppression, on reception, of a broadband radio signal by spatial filtering taking into account different distorted versions of the signal delivered by parallel processing chains, such as for example, the individual processing chains associated with radiating elements of a receiving network antenna.

Certain radioelectric systems such as, for example, the so-called "GPS" satellite navigation system (from the Anglo-Saxon "Global Positioning System"), use wideband spread spectrum radio signals, received with very low power level, easily, from 30 dB to 40 dB below the thermal noise level at the input of a receiver. In a favorable environment, that is to say little noisy, the despreading operation makes it possible to raise the signal-to-noise ratio of the received signal well above 0 dB.Hz, so that the received signals can be exploited from sure way. The spread-band transmission technique is relatively insensitive to intentional interference when the frequency band used is not known to the person seeking to interfere because this frequency band is difficult to spot given the low power density per hertz of the band signal issued. On the other hand, this is no longer the case when the frequency band or bands used are imposed. It is in fact easy to disturb the reception of a spread band transmission signal by drowning it in broadband interference. Indeed, the fact of emitting in all the band occupied by the useful signal, a noise having a power density per hertz clearly higher than that of the useful signal makes it possible to degrade the signal to noise ratio at input of the receiver and this degradation is propagates along the signal processing chain. It therefore suffices that after despreading, it reduces the signal to noise level below a few dB.Hz so that the useful signal becomes difficult to use.

In the case of GPS, the frequency bands used are well known and the strength of the signals received low since they come from transmitters of a few tens of watts on board traveling satellites in orbit distant from around twenty thousand kilometers of the surface of the terrestrial globe These two characteristics of imposed transmission frequency bands and low transmission powers make the reception of GPS signals particularly sensitive to broadband interference since it is sufficient to disturb the reception of GPS signals locally have a noise emitter occupying the GPS frequency bands, in the area concerned, either on the ground or above the ground on board an aircraft, missile or other aircraft. appeared on the market, in 1997, GPS jamming equipment capable, with a relatively low power since of the order of four watts, of disturbing the reception of GPS signals in a very large area whose radius can reach 200 km

To make GPS receivers less vulnerable to broadband interference, various measures have been proposed. The main ones take advantage of the fact that useful signal transmitters, which are navigation satellites, and jammers do not occupy the same geographical locations. act on the radiation pattern of the receiver antenna to favor reception in the directions of origin of the useful signals transmitted by the navigation satellites to the detriment of reception in the directions of the interference signals. A first known measurement consists in using in reception, an antenna with a hemispherical radiation diagram turned towards the sky excluding directions with a low elevation angle. Such an antenna is often of the array type, with several radiating elements placed, on a ground plane, at the vertices of a regular polygon and possibly a central radiating element, and with a space combiner which makes a weighted sum, in amplitude and in phase, signals picked up individually by the different radiating elements of the antenna to obtain a reception channel corresponding to the desired radiation pattern

A second measure complementary to the first, consists, since we are in the presence of a network type reception antenna, in implementing the technique of dynamic reduction of the power of the parasitic signals as opposed to the secondary lobes, known technique under the acronym OLS for "Opposition of Secondary Lobes" (in English CSLC for "Coherent Side Lobe Canceler") This OLS interference suppression technique was originally developed in the context of radars. It consists of dynamically forming, from the signals picked up by several radiating elements, a reception channel corresponding to a radiation diagram presenting one or more several lobes in the directions of the useful signals and holes in the directions of the jammers Its implementation goes through the following main stages

- static formation of several independent reception channels from the signals picked up by the radiating elements used for reception, the independence between reception channels signifying that none of them is reduced to a simple linear combination of the others,

- election among the reception channels formed, of a main reception channel, the others being considered as auxiliary, and

- dynamic formation of a reception channel called "dewormed" by adding, to the signal of the main reception channel, a linear combination of the signals of the auxiliary reception channels weighted in amplitude and in phase using adjusted coefficients dynamically so that the signals of the reception channel called "dewormed" and the auxiliary reception channels are decorrelated

It is shown that the possible number of independent reception channels is less than or equal to that of the distinct radiating elements of the reception antenna and that, in order to be able to eliminate N jammers, it is necessary to have at least N independent auxiliary reception channels and therefore to have when receiving a network antenna comprising at least N + 1 distinct radiating elements

The effectiveness of this OLS interference suppression technique depends on the accuracy and precision of the determination of the amplitude and phase weighting coefficients used to add the signals of the auxiliary reception channels to the signal of the main reception channel Without precautions particular, the use of this OLS interference suppression technique in the context of signal receivers spread band such as GPS receivers, reduces their sensitivity to broadband interference by around 30 dB

One of the reasons put forward for the lack of precision of the mechanisms for determining the amplitude and phase weighting coefficients used to add, to the signal of the main reception channel, the signals of the auxiliary reception channels in the context of the technique of interference suppression OLS is the disruption of these determination mechanisms by narrow band interference which does not necessarily result from a hostile intention but from the increasing congestion of the radio spectrum and is superimposed on broadband interference

It has been proposed to remedy this by using, at the level of each of the main and auxiliary reception channels, before the dynamic formation of the "dewormed" reception channel, narrow-band rejector filters whose central frequencies are adjusted so as to adaptive to those of narrowband interference signals which manifest themselves as peaks exceeding the noise level in the useful band

In fact, there are other possible reasons, including the fact that the mechanisms for determining the amplitude and phase weighting coefficients employ determination methods based implicitly on an identity of the transfer functions of the processing chains encountered on the different reception channels, while the analog nature of the input stages of the processing chains makes it possible to approach this identity of transfer functions only imperfectly, with deviations varying in disparate ways, over time

Indeed, the emission frequency range of most radio signals, this is the case for signals from GPS positioning satellites, is too high to lend itself directly to digital processing. The received radio signal must then undergo, in analog head stages, rejection of out-of-band noise and a drop in frequency, towards a lower frequency range, intermediate frequency band or base band, more accessible to the current technology of analog / digital converters This analog nature of input stages of the processing chains and the wide frequency range to pass for the processing of a spread band signal make it impossible, even starting from an identical conception, to succeed in giving exactly the same transfer function to two distinct processing chains. Thus, there are always, between the transfer functions of the different chains, unwanted differences in amplitudes and phases depending on the frequency considered, which, moreover, evolve in disparate and random ways over time. These inevitable and evolving differences adversely affect the accuracy and precision of the determination of the amplitude and phase weighting coefficients and therefore the effectiveness of the deworming.

The object of the present invention is to improve the efficiency of the O.L.S. in the context of useful wideband signals It aims, more generally, to improve the efficiency of all the processing of broadband signals taking into account different versions of the same signal coming from parallel processing chains supposed to have identical transfer functions.

Its subject is a method of interference suppression, on reception, of a useful radio signal, by spatial filtering taking into account different parasitized versions of the signal delivered by a set of parallel processing chains supposed to have identical transfer functions, remarkable in that it consists in carrying out an equalization, at the levels of the processing chains, before spatial filtering, by individually equipping processing chains with adjustable time equalization filters and with equalization filter adjustment circuits operating from the differences observed between two versions of the same test signal generated locally on reception by a test signal generator, one of the versions traversing the equalized processing chain concerned and the other serving as a reference.

Advantageously, the test signal differs from the useful signal by its clearly higher power which relays the useful signal at the level of ambient noise. Advantageously, the test signal differs from the useful signal by an auxiliary modulation.

Advantageously, the test signal is distinguished from the useful signal by an auxiliary modulation by a pseudo-random binary sequence, the test signal being added to the useful signal in its modulated form during its passage in a processing chain then separated from the useful signal by demodulation before being used by an equalization filter adjustment circuit.

Advantageously, an equalization filter adjustment circuit uses the version of the test signal having traversed the processing chain to be equalized to measure, explicitly, the transfer function of this processing chain to be equalized, compares this transfer function with a reference transfer function taken from the reference version of the test signal and deduced from the observed differences, adjustment values making it possible to attenuate these differences for an equalization filter inserted after the processing chain to be equalized .

Advantageously, an equalization filter adjustment circuit operates in a self-adaptive manner, so as to maximize the. correlation value between the two versions of the test signal, the reference version and the version which has passed through the processing chain to be equalized, including the equalization filter.

Advantageously, the test signal generator provides two versions of the test signal, one in the input frequency band of the processing chains, the other in the output frequency band of the processing chains.

Advantageously, the reference version of the test signal is provided by one of the so-called main processing chains which does not have an adjustable equalization filter and an equalization filter adjustment circuit, and whose effective transfer function is taken for reference transfer function, while all the other so-called auxiliary processing chains are individually provided with an adjustable equalization filter and an equalization filter adjustment circuit

Advantageously, all the processing chains are individually provided with an adjustable equalization filter and an equalization filter adjustment circuit, the reference version of the test signal being taken equal to a weighted linear combination of the versions of the transmitted by the different processing chains and taken upstream of the adjustable equalization filters.

Advantageously, the adjustment of an equalization filter is done during calibration periods during which the useful signal is replaced at the input of the processing chain considered by the test signal.

Advantageously, time slices are reserved for the calibration periods which are simultaneous for the different processing chains, the spatial filtering of deworming being interrupted during these time slices

Advantageously, the calibration periods assigned to the different processing chains are not simultaneous but succeed one after the other, the processing chain under test being temporarily excluded from spatial filtering.

Advantageously, the test signal is a spread band signal which occupies the same frequency band as the useful radio signal and which is obtained by modulation of a carrier with a specific pseudo-random binary sequence.

Advantageously, the test signal is a pure frequency line sweeping the frequency band of the useful radio signal

Advantageously, in the case of parallel processing chains corresponding to reception channels coming from a network antenna. reception with several radiating elements, the test signal is injected into the different processing chains, at the output of the receiving network antenna, by means of multiplexing devices making it possible to connect the inputs of the processing chains either to the antenna of reception, either to the test signal generator.

Advantageously, the adjustment of the equalization filters of the processing chains is done continuously, in the background of the reception of the useful signal, by means of the test signal which is added to the useful signal at the input of the processing chains via a bank of couplers.

Advantageously, the adjustment of the equalization filters of the processing chains is done continuously, in the background of the reception of the useful signal, by means of the test signal which is modulated by an auxiliary modulation using a binary sequence. pseudo-random, which is added to the useful signal at the input of the processing chains by means of a bank of couplers and which is separated from the useful signal by demodulation at the output of the processing chains.

The invention also relates to a suppressor device, on reception of a useful radioelectric signal operating by spatial filtering and implementing the aforementioned method.

Other characteristics and advantages of the invention will emerge from the description below of an embodiment given by way of example. This description will be made with reference to the drawing illustrating, schematically, various examples of interference suppression devices by spatial filtering mounted following a multi-strand reception antenna with several reception channels and provided, in accordance with the invention, with adjustable equalization filters inserted at the output of the processing channels of the reception channels, upstream of the spatial interference suppression filter, and of equalization filter adjustment circuits operating using a locally generated test signal. More specifically, the drawing includes:

- Figure 1 showing an example in which the equalization filter adjustment circuits operate, during calibration periods, by spectral analyzes of a version of the signal test taken at the output of the measurement chain, upstream of the adjustable equalization filter, and of a reference version of the test signal supplied directly by the local generator generating the test signal applied at the input of the processing chains,

FIG. 2 showing an example in which the circuits for adjusting the equalization filters operate, during calibration periods, adaptively, on the basis of a correlation between a version of the test signal taken at the output of the measurement, downstream of the equalization filter, and of a reference version of the test signal supplied directly by the local generator generating the test signal applied at the input of the measurement chain,

FIG. 3 showing an example in which the adjustment circuits of the equalization filters operate, during calibration periods, by spectral analyzes of a version of the test signal taken at the output of the measurement chain, upstream of the filter adjustable equalization, and a reference version of the test signal taken from the output of a measurement chain without an adjustable equalization filter,

FIG. 4 showing an example in which the adjustment circuits of the equalization filters operate, during calibration periods, adaptively, on the basis of a correlation between a version of the test signal taken at the output of the measurement, downstream of the equalization filter, and of a reference version of the test signal taken from the output of a measurement chain lacking an adjustable equalization filter,

FIG. 5 showing an example in which the circuits for adjusting the equalization filters operate, in the background of the reception of the useful signal, by means of the test signal which is added to the useful signal at the input of the processing chains by the intermediary of a bank of couplers, adaptively, from a correlation between a version of the test signal taken at the output of the measurement chain, downstream of the equalization filter, and a reference version of the test signal supplied directly by the local generator generating the test signal applied at the input of the measurement chain,

FIG. 6 showing an example in which the circuits for adjusting the equalization filters operate, in the background of the reception of the useful signal, by means of the test signal which, after having undergone an auxiliary modulation, is added to the useful signal at the input of the processing chains by means of a coupler bank, by spectral analyzes of a version of the test signal taken at the output of the measurement chain, upstream of the adjustable equalization filter, and separated from the useful signal by demodulation , and a reference version of the test signal supplied directly by the local generator generating the test signal applied at the input of the processing chains,

FIG. 7 showing an example in which the circuits for adjusting the equalization filters operate, in the background of the reception of the useful signal, by means of the test signal which, after having undergone an auxiliary modulation, is added to the useful signal at the input of the processing chains by means of a coupler bank, by spectral analyzes of a version of the test signal taken at the output of the measurement chain, upstream of the adjustable equalization filter, and separated from the useful signal by demodulation , and a reference version of the test signal taken from the output of a measurement chain without an adjustable equalization filter, - a FIG. 8 showing an example in which the adjustment circuits of the equalization filters operate, in background of the reception of the useful signal, by means of the test signal which, after having undergone an auxiliary modulation, is added to the useful signal at the input of the processing chains via a ba nc of couplers, adaptively, from a correlation between two versions of the test signal separated from the useful signal by demodulation and sampled, one at the output of a measurement chain to be equalized, downstream of the filter equalization, and the other at the output of a measuring chain without an adjustable equalization filter, and FIG. 9 showing an example in which the circuits for adjusting the equalization filters operate, in the background of the reception of the useful signal, by means of the test signal which, after having undergone an auxiliary modulation, is added to the useful signal directly at the input of each sensor by means of radiation from the central element and by directional coupling to the latter, adaptively, from a correlation between two versions of the test signal separated from the useful signal by demodulation and sampled, l '' at the output of a measurement chain to be equalized, downstream of the equalization filter, and the other at the output of a measurement chain without an adjustable equalization filter

In the different figures, the unchanged elements keep the same indexing

The antiparasitic device by spatial filtering illustrated in FIG. 1, is supposed to process in reception, radioelectric signals coming from the navigation satellites of the GPS II system is placed at the level of the input stages of a GPS receiver, following of the receiving antenna 1, in front of the floors devoted to despreading of GPS signals, to the extraction of the information contained in the despread signals and to the position and speed calculations from the information extracted

The receiving antenna 1 is a network antenna with a hemispherical radiation diagram directed upwards to allow reception of signals from transmitters on board aboard traveling satellites. It is configured to have a sensitivity turned towards the sky and limited to angles. sites greater than a few degrees to be as little as possible influenced by terrestrial jammers, intentional or not As represented, it is formed by a network of seven radiating elements 1 a, 1 b, 1 c,, 1f, 1 g arranged on a substantially horizontal ground plane, six 1 a,, 1f distributed according to the vertices of a regular hexagon and a seventh 1 g placed in the center of the hexagon This representation is theoretical because the receiving network antenna is actually made from the patch antenna technique Other antenna configurations are possible with more or less radiating elements arranged at the vertices of a regular polygon and with a central radiating element or not, the antenna preferably having an omnidirectional radiation pattern in azimuth since the orientation of the receiver. is not necessarily known in advance.

The signals picked up by the radiating elements 1a, ..., 1g of the receiving network antenna are applied to individual identical processing chains, of which only one 2d is shown entirely, ensuring the shaping of the received signal rendering it suitable for purely digital processing. In the usual way, each individual processing chain comprises one or more input stages 20 ensuring a limitation of the out-of-band noise and a descent in frequency of the received signal, from the HF range in which it was transmitted (L bands) to a intermediate frequency range FI more conducive to digitization, an analog / digital converter 21 and an output stage 22 completing the frequency descent and ensuring the passage of the useful signal in baseband.

After having been digitized and placed in baseband by the individual processing chains 2a, ..., 2g, the signals picked up by the various elements 1a, ...., 1g of the antenna 1, which have the in the form of sequences of digital samples with two components, one I called in phase and the other Q called in quadrature, are applied to a spatial interference suppression filter 3.

The spatial interference suppression filter 3 achieves a linear combination of the signals picked up by the different radiating elements of the antenna to create a reception channel favoring the useful signals to the detriment of the interference signals. Its operating principle is based on the assumption that the wanted signals and the interference signals are not picked up in the same directions. It consists in creating, in the radiation diagram corresponding to the reception channel formed at the output of the spatial filter, holes in the directions of the jammers but not in the directions of the useful signals.

The adjustment of the spatial interference suppression filter 3, that is to say the determination of the weighting coefficients of the linear combination of the signals picked up by the reception antenna which it carries out, is done by considering one of the reception channels. applied as input to the spatial filter 3 as a master channel and the others as auxiliary channels, by imposing a unit weighting coefficient on the master channel and by determining the weighting coefficients of the auxiliary channels in order to obtain at the output of the spatial filter a signal without correlation with them. In the examples shown, it is assumed that the master reception channel is that coming from the central element 1 g of the reception antenna 1 but this master channel could come from another radiating element or even result from a particular combination signals picked up by the different radiating elements of the antenna, particular in the sense that it is favorable to the reception of useful signals. The auxiliary reception channels here originate from each of the peripheral radiating elements 1a, ..., 1f of the reception antenna 1 but they could also result from particular combinations of the signals picked up by the different radiating elements of the antenna, particular in meaning that they would be rather favorable to the reception of jammers. The only limitation is that the master and auxiliary reception channels are independent, that is to say that none of them is reduced to a simple linear combination of the others.

This antiparasitic technique will not be detailed further because it is well known in the radar field under the name of anti-jamming as opposed to secondary lobes or anti-jamming O.L.S. in French and C.S.L.C. in English. One of the problems it poses is that it uses methods for determining the coefficients of the spatial interference filter which, in the context of useful broadband signals as are spread band signals, are very sensitive to disparities in the transfer functions of the various individual processing chains supplying the signals of the main and auxiliary reception channels. This leads to the imposition, during the design of independent processing chains, of severe constraints of performance equalization between all the chains to be respected over time, which significantly increase the cost.

To remedy the problem posed by the disparities in the transfer functions of the individual processing chains and thus achieve either to improve the performances of the interference suppression, or to obtain the same level of interference suppression with individual processing chains of lower cost, we interpose at the output of the individual processing chains 2a, ..., 2f, in front of the inputs of the spatial interference suppression filter 3, a bank of automatic equalization circuits coupled to a test signal generator 5 and to a battery 6 of switches making it possible, on command, to replace, at the input of the different individual processing chains 2a, ...., 2g, the signals picked up by the radiating elements 1a, ..., 1g of the antenna of reception 1 by a test signal occupying the same high frequency band, available on an HF output of the test signal generator 5.

The automatic equalization circuit bank contains an automatic equalization circuit per reception channel. Only the one 4d dedicated to the reception channel corresponding to the signal picked up by the radiating element 1d from the reception antenna 1 is visible in FIG. 1. Each automatic equalization circuit 4d includes an adjustable equalization filter 40d inserted in output of the reception channel processing chain 2d to which it is dedicated, upstream of the spatial interference filter 3, a circuit 41 d for determining the coefficients of the adjustable equalization filter 40d, two circuits 42t and 43d analyzers spectrum, one 42t connected as an input, to an IF output, at intermediate frequency, from the test signal generator 5, the other 43d connected as an output from the analog / digital converter 21 d of channel 1d assigned to the automatic circuit d equalization considered 4d, and a subtractor 44d connected at the output of the two spectrum analyzer circuits 42t and 43d providing the difference observed between the two spectra at circuit 41d of determin ation of the coefficients of the equalization filter 40d.

The battery 6 of switches makes it possible to isolate at will the individual processing chains from the influence of the outside world and therefore from the jammers so as to subject them only to the test signal.

The test signal generator 5 generates locally, on an HF output in the high frequency transmission band of the useful signal picked up by the receiver, and on an IF output in the intermediate frequency band occupied by the output signal of the analog converters / digital 21a to 21g of the individual processing chains 2a to 2g, two versions of the same spread band test signal of the same spectral width as the useful signal and of powers adapted to the input sensitivities of the individual processing chains 2a,. .., 2g and 4d equalization circuits. he can implement a carrier modulated in two-phase by a pseudo-random binary sequence.

The test signal, once injected via the battery of switches 6, at the input of the individual processing chains 2a, ...., 2g in place of the signals received consisting of the useful and interference signals picked up by the radiating elements of the reception antenna 1, cross the different stages of the individual processing chains placed upstream of their analog / digital converters. It thus traverses all their analog parts and undergoes, on the part of the individual processing chains 2a, ...., 2g, the same disturbances as the useful signals picked up by the receiving antenna.

The spectrum analyzer circuits 42t, 43d deliver the complex frequency spectra of the signals applied to their inputs. They can operate according to all known techniques of spectral analysis, as well those using the Fourier transformation in all its variants, as those called "high resolution" using an approach by the eigenvalues and eigenvectors of the autocorrelation matrix of the signal or an autoregressive modeling approach. In FIG. 1, the spectrum analyzer circuits 42t, 43d are assumed to operate an FFT, that is to say a rapid Fourier transformation.

One, the spectrum analyzer circuit 42t delivers the complex frequency spectrum of the version of the test signal which occupies the intermediate frequency band output from the processing chain while being directly derived, in digital form, from the output FI of the generator of the test signal 5 while the other 43d delivers the complex frequency spectrum of the test signal after it has passed through the processing chain 2d while it is affected by disturbances due to this processing chain 2d. The complex frequency spectrum delivered by the spectrum analyzer circuit 42t constitutes a reference transfer function model making it possible to take into account the imperfections of the test signal with respect to the flat spectrum sought in the useful band while the complex frequency spectrum delivered by the spectrum analyzer 43d constitutes a measure of the transfer function of the processing chain 2d to imperfections near the test signal. The subtractor 44d subtracts from the transfer function model delivered by the spectrum analyzer circuit 42t, the measured transfer function delivered by the spectrum analyzer 43d to highlight the spectral disturbances to be corrected by the adjustable equalization filter 40d placed at the output of the individual processing chain 2d.

The circuit 41 d for determining the coefficients of the equalization filter performs the conventional calculations for determining the coefficients of a filter to carry out the transfer function delivered by the subtractor

44d. It can be of any known type and operate, for example, by inverting the autocorrelation matrix of the spectral disturbances.

The adjustable equalization filter 40d is a programmable digital filter since it is located, following an individual processing chain 2d, at a level where the signals are available in digital form. It is advantageously a FIR type finite response filter.

For convenience, the spectrum analyzer 42t operating directly on the signal from the test signal generator 5 has been shown as belonging to each automatic equalization circuit 4a, 4b, ..., 4g. It goes without saying that it can exist in only one copy and be shared with all automatic equalization circuits

According to a variant, instead of generating a pseudo-random binary sequence, the test signal generator 5 can generate a pure frequency line of calibrated amplitude which scans the band occupied by the useful signals at the inputs of the individual processing chains 2a, ..., 2g. The spectrum analyzer circuits 42t, 43d then supply complex spectra made up subsequently of complex samples (amplitude-phase) of the signals taken from the outputs of the analog-digital converters of the individual processing chains for different frequency values scanned by the frequency line of the test signal.

FIG. 2 shows a second example of a spatial filtering interference suppression device provided with a bank of automatic equalization circuits. The difference compared to the anti-interference device described relative to FIG. 1 is essentially located, at the level of the design of the automatic equalization circuits 4d ′ and more particularly of the mode of adjustment of the adjustable equalization filter 40d which is no longer made from an explicit measurement of the differences presented by the transfer function of the processing chain with respect to a reference transfer function but according to an adaptive process tending to maximize the value of the correlation existing between the test signal at the output of the adjustable equalization filter 40d and its version coming directly from the test signal generator 5 taken this time on an output BB where the test signal is available in digital form, no longer in intermediate frequency band but in base band, a delay circuit 7 making it possible to correct the delay ΔT1 taken by the test signal from the processing chain due to its longer path and to put the two v in phase correlated ersions of the test signal. Thus, instead of the spectrum analyzers 42t, 43d and the subtractor 44d, there is a simple correlator 45d, one input of which is connected to the output of the adjustable equalization filter 40d and the other following the delay circuit 7 placed on the output BB of the test signal generator 5. The circuit for determining the coefficients of the adjustable equalization filter 41 operates by successive approximations, implementing the conventional techniques based on the gradient algorithm used, for example , in the field of modems to perform a suitable filtering of transmission links.

FIG. 3 shows a third example of a device for interference suppression by spatial filtering provided with a bank of automatic equalization circuits. The difference with respect to the interference suppression device described in relation to FIG. 1 is at the level of the test signal version used for the transfer function model This, instead of coming directly from the test signal generator 5 is extracted from the individual processing chain 2g assigned to the reception channel coming from the central radiating element 1g from the antenna 1 taken as the master channel for the spatial interference filter 3 In this case, the individual processing chain 2g of the master channel is taken for reference and does not have an automatic equalization circuit unlike all the others individual processing chains 2a to 2f. Note that the two versions of the test signal used by an automatic equalization circuit 4d are extracted at the same level from the two individual processing chains, that is, as shown, at the intermediate output of the analog / digital conversion stage, or even at the final output, after stage 22g, 22d of translation of the baseband signal, but upstream of the adjustable equalization filter 40d.

FIG. 4 shows a fourth example of a spatial filtering interference suppression device provided with a bank of automatic equalization circuits. This is distinguished from the previous one by the constitution of the automatic equalization circuits 4d 'which takes up that with correlator 45d of the second example illustrated in Figure 2. As one of the inputs of the correlator 45d must be connected at the output of the adjustable filter equalizer 40d delivering a baseband signal, the other input of the correlator 45d can only be connected at the final output of the individual processing chain 2g while the signal is also in baseband .

In the examples of interference suppression devices by spatial filtering provided with a bank of automatic equalization circuits which have just been described relative to FIGS. 1 to 4, the banks of automatic equalization circuits are adjusted during operations of calibration where the test signal is substituted for the signals of the receiving antenna 1 by an operation of a bank of switches 6. These calibration operations can be carried out simultaneously for all the processing chains, for example when switching on route of the receiver or sequentially after a certain period of time taking into account the drifts of the processing chains. The assignment of different calibration periods for the processing chains allows spatial filtering to continue during the calibration periods, only the processing chain under test being temporarily excluded.

The insertion of calibration periods during the operation of the GPS receiver is advantageously entrusted to an automaton which either regularly imposes calibration periods on the start-up of the receiver and at the end of operating periods of a determined duration chosen as a function of the drift speeds of the characteristics of the individual processing chains, or initializes a calibration period when the receiver is started up and each time the signal to noise ratio of the received signal, measured downstream in the receiver after the despreading operation becomes below a threshold considered as the minimum tolerable for reliable operation of the GPS receiver.

It is also possible to provide interference suppression devices with spatial filtering of bank of automatic equalization circuits which can be adjusted continuously, in the background of the reception of the useful signal.

Figure 5 gives an example derived from that of Figure 2. This fifth example is similar to that of Figure 2 which he uses the design of automatic equalization circuits 4d '. It differs from it only in that the bank 6 of switches 6a to 6g has given way to a bank 8 of couplers 8a to 8g. The test signal generated at high frequency by the test generator 5 is superimposed on the reception signals of the antenna 1 in the different individual processing chains 2a to 2g. At an automatic equalization circuit 4d ′, the useful reception signal superimposed on the test signal at the output of the adjustable equalization filter 40d plays the role of noise in the correlation effected by the correlator 45d, with the version of the test signal coming directly from the test signal generator 5, noise which is eliminated or, at the very least, very weakened if care has been taken to take a pseudo-random binary code orthogonal to the codes for the test signal pseudo-random binaries used by the GPS system.

FIG. 6 shows a sixth example of a spatial filtering interference suppression device with a bank of automatic equalization circuits which can be adjusted continuously, in the background of the reception of the useful signal. This other example is similar to that of FIG. 1, from which it takes up the design of the automatic equalization circuits 4d. It is distinguished from it, however, by different characteristics. As in the previous example in FIG. 5, the bank 6 of switches 4a to 4g has left its place a bench 8 of couplers 8a to 8g but in addition, the test signal available on the HF output of the test signal generator 5 undergoes an auxiliary modulation by a pseudo-random code before being superimposed on the useful reception signal at the input of the individual processing chains 2a to 2g then an auxiliary demodulation at the output of the individual processing chains 2a to 2g before being applied to the spectrum analyzer 43d.

The auxiliary modulation and demodulation of the test signal before and after it has passed through the individual processing chain to be equalized is intended to make it possible to attenuate the useful signal at the input of the spectrum analyzer 43d so that it mainly sees the test signal. They are carried out by means of a local generator 9 of pseudo-random binary code, of a modulator 10 interposed between the HF output of the test signal generator 5 and the bank 8 of couplers 8a to 8g and of a demodulator 11 inserted before the input of the spectrum analyzer 43d of the automatic equalization circuit. The auxiliary modulation is preferably done at a multiple of the bit rate of the test signal and performs an additional band spread which is chosen to accommodate the bandwidth margin on either side of that of the useful signal, adopted for individual treatment chains when they are designed.

FIG. 7 shows a seventh example of a device for interference suppression by spatial filtering provided with a bank of automatic equalization circuits. The difference compared to the interference suppression device described in relation to the previous figure 6 is at the level of the version of the test signal used for the transfer function model. This, instead of coming directly from the test signal generator 5, is extracted from the individual processing chain 2g assigned to the reception channel coming from the central radiating element 1g from the antenna 1 taken as the master channel for the antiparasitic spatial filter 3. As the signal extracted from the individual processing chain 2d subject to equalization, the signal extracted from the individual processing chain 2g taken for reference undergoes a demodulation at 12 intended to bring out the test signal of all the components of the output signal of this individual 2g processing chain. Note that, as previously for the example illustrated in Figure 3, the chain individual processing 2g of the master channel taken for reference does not have an automatic equalization circuit unlike all the other individual processing chains 2a to 2f and only the two versions of the test signal used by an equalization circuit automatic 4d are extracted at the same level from the two individual processing chains, either, as shown, at the intermediate output of the analog / digital conversion stage, or even at the final output, after the stage 22g, 22d of signal translation in baseband, but upstream of the adjustable equalization filter 40d

FIG. 8 shows an eighth example of a device for interference suppression by spatial filtering provided with a bank of automatic equalization circuits. This is distinguished from the preceding illustrated in FIG. 7, by the constitution of the automatic equalization circuits 4d ′. which uses the correlator 45d from the examples illustrated in Figures 2, 4 and 5 As one of the inputs of the correlator 45d must be connected to the output of the adjustable equalization filter 40d delivering a baseband signal, the other input of the correlator 45d can only be connected at the final output of the individual 2g processing chain while the signal is also in baseband

Other variants of interference suppression devices by spatial filtering with bank of automatic equalization circuits are also possible, in particular variants combining the bank of switches and the bank of couplers of the examples previously described. It is also possible to take into account the part of the inhomogeneity faults, in the processing of the reception channels, due to the radiating elements of the reception antenna It is sufficient for this to replace the banks of switches or couplers by a transmission antenna excited by the local generator test signal and placed on the same carrier as the receiving antenna, near or even within the receiving antenna

As shown in the ninth example of an anti-jamming device by spatial filtering illustrated in FIG. 9, one of the radiating elements of the reception antenna, HERE element 1g, can be used as transmitting antenna for the test signal. This radiating element 1 g which retains its role on reception, is excited on transmission, by means of an auxiliary coupler 8 "by the test signal previously subjected to an auxiliary modulation to allow its separation from the useful signal during its reception.

Note that, in the various devices described, the banks 4, 4 ′ of automatic equalization circuits intervene on the signals when they have been digitized. They are therefore produced by digital processing like the spatial filter 3 of interference suppression so that the two functions of equalization by time filtering and of interference suppression by spatial filtering can be carried out by means of the same circuit specialized in signal processing such as a DSP (Digital Signal Processor).

The devices which have just been described have the advantage of making it possible to correct the faults of the analog down-frequency chains of the reception channels of a network antenna by programming digital components which is a much easier operation than intervene in the design of analog stages. We can thus get rid of the often very hard specifications imposed on the HF stages in the field of radio signal sensor networks such as radio navigation signals.

The digital corrections are based on easily modifiable calculation algorithms and, obviously, less subject to technological limitations than those encountered in HF. Finally, from a cost point of view, this shift in complexity from analog to digital allows, for identical performances, to significantly reduce product development costs and leaves open possibilities for optimization by simple update. even in low cost applications.

Claims

1. Antiparasitic method, on reception, of a useful radio signal, by spatial filtering (3) taking into account different parasitized versions of the signal delivered by a set of parallel processing chains (2a, ..., 2f, 2g) supposed to have identical transfer functions characterized in that it consists in carrying out an equalization, at the levels of the processing chains (2a, ..., 2f, 2g), before spatial filtering (3), by equipping individually processing chains (2a, ...., 2g) with adjustable time equalization filters (40d) and with circuits (43d, 42g, 42t, 44d, 41 d, 45d, 41 d ') for adjusting the equalization filter operating on the basis of the differences observed between two versions of the same test signal generated locally, on reception, by a test signal generator (5), one of the versions traversing the equalized processing chain (2d ) and the other for reference.

2. Method according to claim 1, characterized in that the version of the test signal traversing a processing chain to be equalized (2d) is distinguished in the signal delivered by said processing chain to be equalized (2d) by a power clearly greater than those of all other components.

3 Method according to claim 1, characterized in that the version of the test signal traversing a processing chain to be equalized (2d) is distinguished in the signal delivered by said processing chain to be equalized (2d) by an auxiliary modulation.

4. Method according to claim 3, characterized in that said auxiliary modulation is carried out by means of a pseudorandom binary sequence.

5. Method according to claim 1, characterized in that the circuits (43d, 42g, 42g, 44d, 41d) for adjusting the adjustable equalization time filters 40d) use the version of the test signal having traversed the processing chain to equalize (2d) to measure the transfer function of the processing chain to be equalized (2d), compares this measured transfer function with a reference transfer function taken from the reference version of the test signal and deduces from the observed differences, adjustment values for the adjustable equalization filter (40d) allowing these differences to be attenuated.

6. Method according to claim 1, characterized in that the circuits (45d, 41 d ') for adjusting the adjustable time equalization filters (40d) operate in a self-adaptive manner, so as to maximize the correlation value between two versions of the test signal, the reference version and the version that passed through the processing chain to be equalized (2d), including the adjustable temporal equalization filter (40d).

7. Method according to claim 1, characterized in that the local test signal generator (5) supplies two versions of the test signal, one in the input frequency band of the processing chains (2a to 2g) the other in the output frequency band of the processing chains (2a to 2g) to be equalized, the version of the test signal in the output frequency band of the processing chains (2a to 2g) to be equalized serving as a reference.

8. Method according to claim 1, characterized in that one (2g) of the so-called main processing chains, is devoid of adjustable time equalization filter and equalization filter adjustment circuit, and is traversed by a version of the test signal taken for reference to its output from said main processing chain (2g).

9. Method according to claiml, characterized in that all the processing chains (2a, ...., 2g) are individually provided with an adjustable time equalization filter (40d) and with a circuit (43d, 42g , 42t, 44d, 41 d, 45d, 41 d ') of the equalization filter adjustment.

10. Method according to claim 9, characterized in that the reference version of the test signal is taken equal to a weighted linear combination of the versions of the test signal transmitted by the different processing chains (2a to 2g) and taken upstream of the adjustable time equalization filters (40d).

11 Method according to claim 1, characterized in that the adjustment of an adjustable equalization time filter (40d) by an equalization filter adjustment circuit is done during calibration periods during which the useful reception signal is replaced, at the input of the concerned processing chain (2a to 2g) by the test signal.

12. Method according to claim 11, characterized in that time slices are reserved for the calibration periods which are simultaneous for the different processing chains (2a to 2g), the spatial filtering of deworming being interrupted during these time slices.

13. Method according to claim 11, characterized in that the calibration periods assigned to the different processing chains (2a to 2g) are not simultaneous but succeed one another in turn, the processing chain under test being temporarily removed from filtering spatial.

14. The method as claimed in claim 1, characterized in that the test signal is a spread band signal which occupies the same frequency band as the useful radio signal and which is obtained by modulation of a carrier with a pseudo- binary sequence. random.

15. Method according to claim 1, characterized in that the test signal is a pure frequency line sweeping the frequency band of the useful radio signal.

16. Method according to claim 1, characterized in that, in the case of parallel processing chains (1a,, 1g) corresponding to reception channels originating from a reception network antenna (1) with several radiating elements (1 a, ..., 1g), the test signal is injected at the input of the various processing chains (2a,, 2g), at the output of the receiving network antenna, by means of multiplexing devices (6a to 6g) allowing the inputs of the processing chains to be connected either to the reception antenna (1) or to the local test signal generator (5).

17. Method according to claim 1, characterized in that, in the case of parallel processing chains (1 a,, 1g) corresponding to reception channels originating from a reception network antenna (1) with several radiating elements ( 1a 1 g), the test signal is injected, at the input of the various processing chains (2a,, 2g), at the output of the receiving network antenna, by means of coupling devices (8a to 8g) making it possible to add the test signal to the useful reception signal.

18. Method according to claim 1, characterized in that, in the case of parallel processing chains (2a, ...., 2f) corresponding to reception channels coming from a reception network antenna (1 ') to several radiating elements (1a, ...., 1f), the test signal is injected into the different processing chains (2a 2f) by a transmitting antenna mounted on the same carrier as the receiving antenna (2) .

19. Device for suppressing, on reception, a useful radio signal, comprising a set of parallel processing chains (2a to 2g) assumed to have identical transfer functions and providing different parasitized versions of the useful radio signal, and a filter interference suppressor (3) performing a weighted linear combination of the output signals of the parallel processing chains, characterized in that it comprises, a test signal generator (5), injection circuits (6, 8) of the test signal supplied by the local test signal generator (5), at the inputs of the parallel processing chains and, at the outputs of the processing chains (2a to 2g), before the spatial interference suppression filter (3) , a bank of automatic equalization circuits (4d, 4d ') individually equipping the processing chains (2a to 2g) and operating from said test signal.

20. Interference suppression device according to claim 19, characterized in that an automatic equalization circuit (4d) comprises at least a first spectral analysis circuit (43d) operating on the signal test at its output from the processing chain (2d) assigned to the automatic equalization circuit (4d) considered, a second spectral analysis circuit (42t, 42g) operating on a version of the test signal taken for reference, a circuit subtractor (44d) subtracting the spectrum measured by the first spectral analysis circuit (43d) from that delivered by the second spectral analysis circuit (42t, 42g), an adjustable time filter (40d) of equalization interposed between the output of the processing chain (2d) assigned to the equalization circuit and the spatial filter (3), and a circuit (41 d) for determining the coefficients of the adjustable time equalization filter (40d) operating on the signal delivered by the subtractor circuit (44d).

21. Device according to claim 20, characterized in that the local test signal generator has at least two outputs, a first output (HF) on which a first version of the test signal is available in the frequency band input of the processing chains (2a to 2g) and a second output (FI, BB) on which a second version of the test signal is available in the output frequency band of the processing chains (2a to 2g) to be equalized , the version of the test signal in the output frequency band of the processing chains (2a to 2g) to be equalized serving as a reference.

22. Device according to claim 20, characterized in that the bank of automatic equalization circuits (4d) does not include an automatic equalization circuit for one (2g) of the so-called main processing chains which delivers the version taken for reference of the test signal.

23. Device according to claim 20, characterized in that the bank of automatic equalization circuits (4d) comprises an automatic equalization circuit for each of the processing chains (2a to 2g).

24. Device according to claim 23, characterized in that the automatic equalization circuits (4d) receive for reference version of the test signal, a version of the test signal established from a weighted linear combination of the test signals issued by different processing chains (2a to 2g), upstream of their adjustable time equalization filters (40d).

25. Device according to claim 19, characterized in that an automatic equalization device (4d ') comprises an adjustable time filter

(40d) of equalization inserted between the output of the processing chain (2d) assigned to the equalization circuit and the spatial filter (3), a correlator (45d) performing a correlation between a reference version of the test signal and a version of the test signal from the adjustable time equalization filter, and an adjustment circuit (41 d ') of the equalization filter operating in a self-adapting manner, so as to maximize the correlation value supplied by the correlator (45d) .

26. Device according to claim 25, characterized in that the local test signal generator has at least two outputs, a first output (HF) on which a first version of the test signal is available in the frequency band input of the processing chains (2a to 2g) and a second output (BB) on which a second version of the test signal is available in the output frequency band of the adjustable equalization time filters (40d), the version of the test signal in the output frequency band of the time equalization filters (40d) serving as a reference.

27. Device according to claim 25, characterized in that the bank of automatic equalization circuits (4d ') does not include an automatic equalization circuit for one (2g) of the so-called main processing chains which deliver the version taken for reference of the test signal.

28. Device according to claim 25, characterized in that the bank of automatic equalization circuits (4d ') comprises an automatic equalization circuit for each of the treatment chains (2a to 2g).

29. Device according to claim 28, characterized in that the automatic equalization circuits (4d ') receive for reference version of the test signal, a version of the test signal established from a weighted linear combination of the test signals delivered by the different processing chains (2a to 2g), upstream from their adjustable time equalization filters (40d).

30. Device according to claim 19, characterized in that it comprises a bank (6) of switches making it possible to replace, during calibration periods, the useful signals at the inputs of the processing chains with the test signal.

31. Device according to claim 19, characterized in that it comprises a bench (8) of couplers making it possible to add, to the useful reception signal, at the input of the processing chains (2a to 2g), the signal from the local generator test (5).

32. Device according to claim 19, with parallel processing chains corresponding to reception channels coming from a reception network antenna with several radiating elements, characterized in that it comprises a transmitting antenna excited by the local signal generator test (5) and mounted on the same carrier as the receiving network antenna.

33. Device according to claim 19, characterized in that the test signal generator is a pseudo-random binary sequence generator.

34. Device according to claim 19, characterized in that the test signal generator is an isolated line generator sweeping in frequency the band of the useful signal.

35. Device according to one of claims 30, 32, characterized in that it comprises an automaton introducing a calibration period at each of its starts.

36. Device according to one of claims 30, 32, characterized in that it comprises an automaton introducing a calibration period after a determined period of operation.

37. Device according to one of claims 30, 32, characterized in that it comprises an automaton introducing a calibration period each time the signal to noise ratio downstream of the spatial filter falls below a threshold value predetermined.

38. Device according to claim 19, characterized in that it is applied to a receiver of a satellite navigation system.

39. Device according to claim 32, characterized in that said transmitting antenna consists of one (1g) of the radiating elements of the receiving antenna, by means of the installation of a directional coupler (8 ").