FR2889770A1 - Parasitic signal suppressing method for phased antenna array, involves constructing directional parasitic vector relative to direction of parasitic sources, and calculating projection operator transforming directional vectors to null vector - Google Patents

Abstract

The method involves generating signals from antenna elements which reflect the reception of an interested signal and parasites in the elements. A direction of parasitic sources received with the interested signal is determined in adaptive manner. A directional parasitic vector relative to the direction of the parasitic sources is constructed. A projection operator transforming the directional vectors into a null vector is calculated and applied to the signals. The interested signal is re-established by compensating a distortion in the interested signal by the application of the operator. An independent claim is also included for an apparatus for suppressing parasites in an interested signal.

Description

The present invention relates to methods and apparatus for realizing

  adaptive way the algebraic suppression of spurious signals and, more

  particularly, techniques for adaptively determining the angular position of spurious signals and selectively removing spurious signals without distortion of a received interesting signal.

  The phased array antennas or electronic scanning antennas consist of an array of individual antenna elements arranged in a particular manner for transmitting and receiving in cooperation with each other directed electromagnetic energy beams. A network antenna beam or beam pattern, which usually includes a main lobe and side lobes, defines the angular dependence of network gain. The shape and direction of a network antenna radiation pattern are determined by the relative phases and amplitudes applied at the individual antenna elements that make up the array, through a process known as beamforming. . By adjusting the relative phases of the antenna elements, it is possible to point the main lobe of the antenna radiation pattern over a range of different directions to transmit a signal in a selected direction or to receive a signal from a direction special. At the moment of receiving a signal, the received power is optimized by pointing the main lobe of the antenna array radiation pattern in the direction of the source of a signal of interest.

  Ideally, the signal in question is aligned with the reference axis of the antenna radiation pattern to optimize the intensity of the received signal. However, parasitic signals present in the side lobes or in the main lobe of the antenna radiation pattern may also be received at the same time as the signal of interest. Adaptive interference suppression systems for phased array antennas have been in use since the 1960s (see, for example, "Adaptive Networks" by S. Applebaum, Proceedings of the Syracuse University Research Group, SPL TR 66-1, August 1966, and "Adaptive Antenna Systems" by B. Widrow et al., IEEE Report, Vol 55, December 1967). There are many variations of these systems, but all involve the following process. At each sampling, a beam forming operation is performed on all network outputs. The network outputs constitute a vector of complex numbers which represents an interesting signal and possibly parasites from one or more disturbing sources, such as jamming signals, for example. The weighting values used for the beamforming operation, values that are calculated during an adaptive process, give a radiation pattern that is deliberately distorted, so that null values are applied to the chart at the level of the beam. angles corresponding to the directions of the disturbing sources. As Hudson explains in "Principles of Adaptive Networks", pages 39 to 48, Peter Peregrinus, London, 1981, working with the weighting values obtained by an optimized adaptive process amounts to performing a beamforming operation in which the Beam pointing vectors were projected in a space orthogonal to the directions of interference.

  Orthogonality is manifested by the presence of zeros in the radiation pattern.

  The adaptive cancellation / shaping (ABF) operation may be considered as a process in which a radiation pattern is formed representing a signal plus parasites, from which a radiation pattern representing only the parasites is subtracted, as is suggested by Hudson and Monzingo in "Introduction to Adaptive Networks", Wiley-Interscience, New York, 1980. Thus, when the angular separation between signal and parasites becomes less than the width of the main lobe, the parasite and signal bundles begin. to overlap, and the suppression of parasites can not be done without attenuating and also deform the signal of interest, this attenuation and this deformation accentuating and worsening with the decrease of the angular separation signal / parasites. The parasites present in the main beam lobe of an antenna therefore pose a serious problem for ABF systems, and the accuracy of the suppression of the parasites is limited by the definition of the antenna network. In addition, since ABF systems suppress noise by deforming network radiation patterns, distortions of the output timing function representing the signal are inevitable. In addition, since the beam forming process used is irreversible, it is not possible to correct the signal distortion. It would therefore be desirable to have an improved noise suppression technique, particularly for sources of noise in the main beam, while distorting the signal of interest.

  According to a first aspect of the present invention, there is provided a method of suppressing noise of an interesting signal received by a phased array antenna comprising multiple antenna elements, characterized in that it comprises the steps which comprises: (a) respectively generating a plurality of antenna element signals which reflect the reception of the interesting signal and noise at the multiple antenna elements; (b) adaptively determining the direction of one or more sources of noise received at the same time as the signal of interest; (c) constructing, for each source of noise, a directional noise vector corresponding to the direction of the noise source; (d) calculating a projection operator that transforms each directional noise vector into a null vector; (e) applying the projection operator to the antenna element signals to eliminate the parasites therefrom; and (f) restoring the signal of interest by compensating individually in the antenna element signals for an interesting signal distortion due to the application of the projection operator.

  The projection operator may be a matrix of MxM A of rank M-k, M being the number of antenna elements of the antenna array, where k is the number of noise sources to be eliminated and k <M. Matrix A is computed to satisfy the expression Ain = 0, for n = 1 to k, where in represents the directional noise vectors. The application of the projection operator eliminates spurious signals, but also distorts the interesting signal. It is, however, possible to exactly restore the signal of interest by dividing each transformed antenna element signal by a corresponding factor which can be determined from the calculations necessary to generate the projection operator.

  The adaptive process of determining the sources of noise sources can be essentially the same type of adaptive process as used in conventional adaptive beamforming systems in which weighting values are obtained from the received signal. However, unlike adaptive beamforming systems, these weighting values are not used during a beamforming operation in the adaptive algebraic noise suppression technique of the present invention. In contrast, the weighting values are used in an algebraic operation to eliminate noise from the directions specified by the weighting values, without distorting the signal of interest. The invention essentially utilizes the capabilities of adaptive beamforming techniques to determine the angles of parasite sources, but, instead of performing a beamforming operation, utilizes this information in a more efficient pest removal system. performance.

  The adaptive algebraic noise suppression process (AAIC) is implemented by a purely deterministic algebraic operation directly applied to the antenna element signals that constitute at each sampling a vector of complex numbers, to selectively eliminate the parasite level. of each network element. Beam shaping is not performed in the AAIC process, and no network features, such as a definition and radiation diagrams, come into play. Therefore, an interesting undeformed signal can be retrieved. . Furthermore, since the AAIC technique operates independently of network radiation patterns and a definition, there is no fundamental limit to a minimum angular separation (except for zero) between a source of noise. and an interesting signal that can be observed without being deformed by radio interference or by the process of eliminating these parasites.

  According to a second aspect of the present invention, there is also provided an apparatus for suppressing noise from an interesting signal, characterized in that it comprises: a phased array antenna comprising multiple antenna elements which respectively generate several antenna element signals reflecting reception of the interesting signal and interference at the antenna elements; and a processor or means adapted to adaptively determine the direction of one or more sources of noise received at the same time as the signal of interest; to construct, for each source of noise, a directional noise vector corresponding to the direction of the source of noise; to calculate a projection operator that transforms each directional noise vector into a null vector; to apply the projection operator to the antenna element signals to eliminate noise from them; and to restore the interesting signal by compensating individually in the antenna element signals for an interesting signal distortion due to the application of the projection operator.

  The foregoing, as well as other objects, features and advantages of the present invention will become more clearly apparent from the following detailed description of preferred embodiments given by way of non-limiting example with reference to the accompanying drawings in which all same reference numerals are used to denote similar elements, and wherein: Fig. 1 is a flowchart of an adaptive technique for algebraically removing parasites according to an exemplary embodiment of the present invention; Fig. 2 is a block diagram of a signal processing device configured to receive a signal through a phased array antenna and to perform adaptive algebraic noise suppression according to an exemplary embodiment. of the present invention; Fig. 3 is a graphical representation of a signal having side-lobe pests and noise suppression resulting from the execution of a conventional adaptive beam shaping / cancellation operation; Fig. 4 is a graphical representation of a signal having pests in the main lobe of the antenna beam and the elimination of pests resulting from performing a conventional adaptive beamforming / cancellation operation; Fig. 5 is a graphical representation of a signal having parasites in the main lobe of the antenna beam and the elimination of noise resulting from performing adaptive algebraic noise suppression in accordance with an exemplary embodiment. of the present invention; Fig. 6 is a graphical representation of the ratio of the signal power resulting from the adaptive algebraic noise suppression to the signal power resulting from the adaptive beamforming / cancellation as a function of the offset angle of the eliminated noise with respect to the signal reference axis; Fig. 7 is a graphical representation of the ratio of signal power resulting from adaptive algebraic noise suppression to signal power resulting from adaptive beamforming / cancellation for selected cases of several spurious signals; and Fig. 8 is a graphical representation of the relative signal power loss resulting from adaptive algebraic noise suppression and the relative signal power loss resulting from adaptive beam shaping / cancellation for selected cases of spurious signals.

  As previously indicated, conventional beam-shaping techniques are limited in their ability to remove closely spaced sources of noise from a signal source of interest. When a beam formed on the noise source is subtracted from the signal beam, a zero value is generated in the signal radiation pattern; however, the cancellation is not selective and affects both the reception of an interesting signal and the noise. The more the signal and the noise will be angularly close, the greater the elimination of the signal at the same time as parasites by the process. This limits the minimum signal / parasitic separation that can effectively be removed by beam shaping techniques. The limitation is mainly due to the beamwidth of the network radiation pattern, which is determined by the physical characteristics of the network antenna. Therefore, in the case of using conventional network beam forming techniques, the noise elimination capability in the near vicinity of a signal of interest is considerably limited by the physical characteristics of the network antenna.

  Nevertheless, an advantageous characteristic of conventional adaptive beamforming techniques is that they essentially determine the angles at which pests are present. Specifically, an adaptive process generates weighting values that are used during a beam shaping operation to deform the radiation pattern by applying zero values at the angles corresponding to the directions of the noise sources. In adaptive beamforming, the operating mode using the weighting values is equivalent to operating with a set of beam pointing vectors that have been projected in the zero space of a beam. matrix orthogonal to the directions of the parasites. In other words, the angles corresponding to the directions of the parasites can be determined directly from the weighting values generated by an optimized process of adaptive beamforming.

  In accordance with the invention, an adaptive process, such as that used for conventional adaptive beamforming operations, is employed to generate information (for example, the weighting values mentioned above) that specify the angles corresponding to the directions. from sources of parasites. However, unlike adaptive beamforming, these weighting values are not used for a beam shaping operation. On the contrary, they are used in an algebraic operation to eliminate noise from the angles specified by the weighting values, without distorting the signal of interest. In fact, the invention makes use of the possibilities of adaptive beamforming techniques to determine the angles of the noise sources, but instead of performing a beamforming operation, it uses this information in a more sophisticated system. pest elimination.

  Specifically, an adaptive algebraic parasite suppression (AAIC) technique can be used to eliminate parasites. This technique fundamentally differs from adaptive beamforming (ABF) systems in many ways. In the first place, the elimination of parasites by the AAIC technique is accomplished by a purely deterministic algebraic operation directly applied to the network outputs which constitute at each sampling a vector of complex numbers. Beam formation plays no part in the process, and no network features, such as a definition or radiation patterns, come into play. Second, with the AAIC technique, parasites are selectively eliminated at the same time. level of each element of the network antenna. Therefore, instead of providing a single distorted output representing the signal of interest, as in the case of ABF techniques, with the AAIC technique, the network radiation patterns after parasite removal and signal recovery are identical to what they are. would be in the absence of parasites. Furthermore, since the AAIC technique operates independently of network radiation patterns and definition, there is no fundamental limit on the minimum angular separation (except for zero) between a source of noise. and an interesting signal that can be observed without being deformed by radio interference or by the process of elimination. The practical limit with respect to this capacity is determined by the receiver noise, as will be described later.

  In the present context, the terms "antenna beam (x), or simply" beam (x) ", refer generally to transmitted energy that is concentrated in a certain direction or sensitivity or receptivity of the antenna to the antenna. signals from particular directions The antenna beam is formed by a directional antenna and is directed at a particular angular area according to the radiation pattern of the antenna (i.e., the radiated field strength depending on the angle), which results in a directed signal The antenna beam can be fixed in a particular direction or electronically or mechanically pointed over a range of directions.

  Usually, it is possible to electronically point the antenna beam by controlling the relative phases of the signals transmitted by the individual antenna elements in a network antenna.

  The flowchart of FIG. 1 represents the steps performed during the adaptive algebraic noise suppression process (AAIC) in accordance with an embodiment of the present invention. In step 102, each antenna element M that collectively constitutes a phased array antenna receives a corresponding radiation signal (e.g., an RF signal). The radiation signals may comprise an interesting signal from an angle substantially centered on the reference axis of the main lobe of the receiving antenna radiation pattern, as well as spurious signals arriving at different angles which may be located in the main lobe or side lobes of the antenna radiation pattern. Each antenna element of the network separately receives a signal having a particular amplitude and phase; therefore, the received radiation signal is actually a set of M signals received by the individual antenna elements.

  Next, step 104 is to adaptively determine information (e.g., weighting values) from the received signal, the information indicating angles corresponding to the direction of spurious signals. The AAIC technique uses essentially the same adaptive process as ABF techniques. For example, a least squares (LMS) known adaptive algorithm can be used to obtain convergence on optimal weighting values. However, a significant difference between AAIC and ABF techniques is the use of weighting values generated during the process. As mentioned above, in the ABF technique, the way to work with the weighting values is to work with a set of beam pointing vectors that have been projected in the zero space of a matrix. In the case of the AAIC technique, the projection operator defined by the weighting values is applied not to beam pointing vectors, but to the observable elements, that is to say to the network outputs which, at each sampling, represent a vector of complex numbers. The application of the projection operator in the AAIC technique transforms the vectors of parasites into a zero vector. The vectors of parasites are annihilated by the transformation in space of zeros realized by the projection operator. In the case of a network of M elements, M-1 noise sources can be annihilated by a single projection operation, more precisely a matrix multiplication.

  Therefore, in step 106, for a network of M elements, a vector with Mx1 columns i corresponding to the direction of a source of noise can be constructed from the information provided by the training formation weighting values. beam: (1) where the phases a, have the form En = 27 [fna, in which T is the delay between elements corresponding to the incident parasitic wavefront. This information is given directly by the weighting values determined from the adaptive process, since the Fourier transform of the weighting vector has zero values for the angles that correspond to the directions of the parasites. Thus, by converging on the optimal weighting values, the adaptive process provides the information necessary to generate the projection operator AAIC.

  The next step 108 of the AAIC process consists in constructing a matrix A such that for a vector of parasites i, the matrix multiplication of the matrix A and the vector i is equal to zero: Ai = O (2) The matrix A is essentially a mathematical tool that can be used to eliminate up to M-1 potential sources of interference received by a network antenna having M network elements. In general, for equation (2) to be satisfied by a non-trivial solution, i 0, it is necessary that A has a rank deficit; that is to say, it can not have an inverse (if the matrix of MxM A is of maximum rank, there exists only the trivial solution i = 0). In the case of k <M jamming signals, a matrix A having a rank deficit, of rank Mk, satisfying the following expression is sought: Ai = 0, for n = 1, 2, ..., k (3 In which the vectors in are the Mx1 directional noise vectors having the form of equation (1) which populate the space of zeros k dimensional of A. The matrix A is constructed using a theorem derived from of the matrix algebra, which states that if a Hermitian matrix of MxM K has an eigenvalue X, repeated k times, the matrix XI - K is of rank Mk, where I is the identity matrix of MxM. If we consider the matrix of MxM K written in the dyadic form of k KE VmVm m = 1 (4) expressed as a sum of external products of vectors, where H denotes the Hermitian conjugate and the vm represent a set of column vectors orthogonal to M elements, the matrix K is Hermitian and has a self-value repeated k times equal to unity (the other roots are all equal to zero). Therefore, the matrix A is given by: A = I K (5) In the case of k parasite vectors, I - K has a space of zeros k dimensional. This means that there are k vectors of parasites in different from zero, where n = 1,2, ... k, such that: Ai = (IK) i = 0 (6) The vectors vm are determined by the resolution of the set of equations: k Ain = (I-EVn, vm) in = 0 n = 1,2, ..., k (7) md or equivalently k E (vn ,, in) vn = in n = 1,2, ..., k (8) m = 1 As previously indicated, the knowledge of the angular positions of the parasites makes it possible to determine in all the k directional noise vectors in equations (7) and ( 8). Once the vectors vm are determined, the matrix A is calculated from equations (4) and (5).

  Thus, by searching for the matrix A, we arrive at a Gram-Schmidt orthogonalization procedure. This illustrates another major difference between the described AAIC technique and the beam shaping / cancellation systems. Here, the Gram-Schmidt procedure is applied to produce an orthonormal coordinate system as a base set to represent the set of selected directional vectors. In beamforming / cancellation systems, the Gram-Schmidt procedure is used to orthogonalize: Beam pointing vectors versus parasitic vectors. In the context of the present invention, the AAIC operators generated by the Gram-Schmidt procedure are applied to different objects for totally different purposes.

  Once the matrix A has been calculated, it can be applied to the outputs of the antenna elements to eliminate noise from the k directions of the interfering signals (step 110 in Fig. 1). More specifically, the interesting signal s arriving at the level of the M elements of the network can be represented by the column vector Mx1: (9) ei 41-in where the phases q have the form cpn = 2nfnr in which i represents the delay between elements for the incident wavefront of the interesting signal. When k parasitic signals In arrive at the same time as the signal interesting s, the received signal plus the parasites (that is to say the outputs of the M antenna elements) is given by the sum of the vectors: k S + E1n = S + 11 +12 + ... + 1k = S + 1 ,,, n = 1, 2, ..., k (10) n = 1 Since the matrix A eliminates each of the k parasitic signals, it follows that: k A (s + Eia) = As = s' n = 1 In other words, the multiplication of the matrix of MxM A by a vector Mx1 of the outputs of the M antenna elements gives a vector Mx1 of modified antenna element outputs in which the spurious signals have been eliminated. This projection operation carried out using the matrix A, which annihilates the parasites, also applies to the signal a transformation which must be eliminated. Since matrix A is singular by construction, this can not be accomplished by inverting the matrix.

  However, the signal transformed at the mth element level of the network can be written in the form of: M-1 M-1 sm = ASm = (IK) Sm -E Km'nei2d (tr,) = ei2nd (tz, n) IE Km'n n = 0 n = 0 ei2nf (tm-t) (12)

L J

  where K, n, n are the elements of the matrix K, and Tm are the inter-element delays for the interesting angle of detection (i.e., the signal of interest). Thus, while the projection operation performed by the matrix A eliminates the spurious signals, the transformation that this process imposes on the interesting signal amounts to multiplying the signal, at the level of the mth element sm, by the given complex number T (m). by: M-1 K ei2 (rn, -r) m, n T (m (13) n = 0 Thus, after the transformation due to the matrix A, the modified signal s'm at the level of the mth element is equal to T (m) * sm, sm representing the original interesting signal received at the mth element.

  Therefore, to restore the interesting signal in its initial state after the elimination of the parasites, the modified output of each element s'm is divided by the corresponding factor T (m) to give the interesting original signal sm without parasites (step 112 in Figure 1). It should be noted that K, and therefore T (m), is always an exactly known quantity. For any value of i, therefore, the signal can always be exactly restored irrespective of errors, if any, regarding adaptively determined parasitic positions. However, since the magnitudes of the T factors are always less than unity, any noise present in the system will be amplified. Since the magnitudes of the factors T diminish progressively with the approximation of the signal and the noise in terms of angular proximity, the noise imposes a practical limit on the minimum angular separation between the signal and the sources of parasites which can be eliminated satisfactorily. .

  A very simple example based on a two-element network will now be described to help illustrate the AAIC process. Suppose that only one interesting signal and one spurious signal are received from two respective directions. The signal and the parasite can be represented at the level of the two-element network by the following vector equations (14) and (15): (14) (15) in which the directions of the signal and the parasite are respectively represented by the angles 0 and (p, and each have the form 27tfi, and where T represents the propagation delay between the first and second network elements in these examples (Te representing the delay corresponding to the signal wavefront and t, p representing the delay corresponding to the parasitic wave front).

  A matrix A with a rank deficit is constructed, unit rank in this example and whose zero space contains the vector i, matrix that is given by the equation: 1 1 A = -e '1 and the multiplication of s + i by A gives: A (s -I- i) = -_; _, or T = 2 -e-, o 1 1 1 e; a + [e, 4,11 IT 0 if 1 + eT * 0 s2 -e '(16) (17) Thus, the parasite i is eliminated and the division of s1' by T and s2 'by T * makes it possible to invert the transformation applied by A to the signal, which gives the desired signal vector: Thanks to the AAIC technique, the sources of parasites are selectively eliminated at each network element by the transformation of directional noise vectors into zero vectors. The conventional process of beam shaping plays no role. After the elimination of the parasites by the AAIC technique, beam shaping, if performed, gives an undistorted signal, as if the parasites had not existed. Signals of interest can be recovered accurately. It should be noted, however, that when suppressing the effect of the transformation by A on s by the elimination of complex multiplicative factors, [1 ei (") i_ e- (eq>) 1 T = J and T * = 2 2 1 (19) the denominators of equation (18) gradually diminish in magnitude as the signal and parasitic angular approximation approaches .. As has been indicated previously, this will amplify a noise or other parasites Therefore, the minimum allowable but nonzero angular separation between an interesting signal and a potential source of noise that is to be eliminated is a function of the noise characteristics of the receiver. other than considerations relating to the (18) signal-to-noise ratio with respect to the minimum angular separation.

  A significant difference between conventional systems based on beam shaping techniques and the AAIC technique described herein is in the objects to which a transformation is applied. In conventional techniques, a transformation operation is applied to the beam pointing vectors for projection into a sub-space orthogonal to the parasites. The transformed vectors are then applied to the more parasitic signal assembly during a traditional beamforming operation in order to generate a zero in the direction of the parasites, whereby the parasites will be reduced, the interesting signals angularly close to the parasites , ie nominally within a bandwidth, also being affected by the zero. Thanks to the AAIC technique of the present invention, the transformation is applied directly to the network output s + i, which selectively eliminates the parasite components directly from each network element to leave only the signal vector transformed but recoverable. , without physical network issues, such as limitations imposed by an angle definition and a network beamwidth, or beam pointing vectors come into play.

  The block diagram of Figure 2 conceptually represents the functional modules of a system 200 for applying the AAIC technique described above to a signal received at the same time as parasites at a network antenna. Figure 2 is a conceptual diagram showing main functional units and a general architecture, and does not necessarily represent physical relationships. Signals from one or more sources (e.g., an interesting signal and spurious signals) are received by a network antenna 202 where signals received by individual network elements can be phase shifted and amplitude adjusted according to of a radiation pattern of the network antenna. An output of each of the respective network elements is transmitted to a receiver 204 which collects and organizes the signal information for presentation to a processor 206 for processing. The processor 206 receives and processes the network outputs according to the AAIC system described above. Depending on the nature of the array antenna, phase and amplitude adjustments may be made by the receiver 204 and not by each of the respective antenna elements of the network antenna 202.

  The processor 206 may be in the form of a single processor or of several different processors performing different functions. For example, the processor 206 may be in the form of any combination of hardware and software that can be statically and / or dynamically configured to implement the AAIC technique described above.

  To compare the performance of the AAIC technique with that of conventional adaptive beamforming / cancellation systems, two fundamental points to consider are signal distortion and signal attenuation.

  For these two points, there are two interesting angular zones, that is to say the area of the side lobes and the zone of the main lobe. It is generally accepted that ABF systems give satisfactory results with regard to the elimination of side lobe parasites, but it is also readily acknowledged that ABF systems have serious difficulties in treating parasites in the main lobe. These two issues are taken into consideration by the following performance analysis for the two angular ranges, which reveals the much higher performance of the AAIC technique compared to ABF systems with respect to the main lobe parasites.

  In order to compare the extent to which the ABF and AAIC processes distort the signal of interest, a network of four elements is considered, an array in which a half-wavelength interval is provided between the elements, and the Resulting radiation patterns after parasite removal by both methods are compared. As mentioned above, beam formation plays no role in the elimination of parasites by the AAIC technique. Figure 3 shows a curve of the radiation pattern for the signal alone at 12 and a curve of the normalized radiation pattern for the signal plus three interference signals at 85 and -80. The ABF technique was applied by a beamforming operation during which the beamforming vectors were projected into a space orthogonal to the pests. As explained in Hudson, supra, this gives optimal performance with respect to ABF systems. As can be seen from the resultant radiation pattern curve shown in Figure 3, the parasites have been eliminated, but the diagram differs considerably from that corresponding to the signal alone and, in fact, provides an erroneous indication. of the signal position, which proves the distortion introduced into the signal by the ABF process.

  In the example shown in Figure 3, the parasites are located in the side lobes, far from the main lobe. FIG. 4 shows the results of an ABF process in the case of three interference signals, the first two of which are located at 11 and 13, at 1 distance from each side of the signal, and the third at -10, and which are so all three inside the main lobe. As can be seen in Figure 4, the radiation pattern after the application of the ABF process is extremely distorted to the point of being unusable.

  Figure 5 shows the results corresponding to a case of parasites similar to that of Figure 4, but in which the AAIC technique is used to eliminate parasites, instead of a conventional ABF process. The curve corresponding to the signal plus the parasites is therefore identical to that of FIG. 4, but the diagram after parasite elimination and signal recovery by the AAIC technique described above is identical to that of the signal alone. This is proof of the absence of distortion imposed on the signal, which is of course a significant difference compared to the ABF system. As previously indicated, this result is possible because the AAIC process provides separate signal outputs for each network element, thereby eliminating the effect of the transformation resulting from the projection operation on the network. signal.

  Figures 3 to 5 show results in terms of radiation patterns. In FIGS. 6 to 8, the relative effects of signal attenuation due to the AAIC and ABF processes are compared as a function of the angular offset between the signal of interest (located at the reference axis) and one or more signals. parasites. In each case, the parasites are eliminated, and the combined signal power provided by the four elements of the antenna is calculated. Since the importance of attenuation of the signal generated by the two methods is compared, the signal power for the AAIC technique used to generate the curves of Figures 6 to 8 is that resulting from the transformation process ( that is, using the matrix A) before the signal power is restored by dividing by the corresponding factor T. In practice, of course, the loss of signal power resulting from the AAIC process is recovered, albeit with a corresponding increase in noise.

  Fig. 6 shows the ratio of the output signal power AAIC to the output signal power ABF as a function of the offset angle with respect to the axis of preference for a single interfering signal. It can be seen that in the area of side lobes, beyond 20 in this example, the signal powers AAIC and ABF are equal, as expected. Once inside the main lobe, however, the additional signal power of AAIC over ABF becomes extremely important. This difference even increases, by several orders of magnitude, in the case of several parasites (for example, three in the present case), as can be seen in FIG.

  Finally, Figure 8 shows, for two interfering signals, the normalized AAIC and ABF signal power levels relative to the power levels attained when the interferences are in the side lobes and the attenuation of the signal power is minimal. . Consistent with the results of FIG. 7, the additional signal attenuation of ABF relative to AAIC becomes considerable for parasitic signals of the main beam and increases progressively as the parasites approach the reference.

  It will be understood that the embodiments described above and illustrated in the drawings represent only a few of the many ways in which the disclosed AAIC methods and apparatus can be implemented and applied. The present invention is not limited to the particular applications described herein, but can be used in many ways. For example, the limitations of conventional adaptive beamforming techniques in terms of the inability to eliminate spurious signals in the main beam without distorting the signal of interest are easily overcome by the described methods and apparatus based on advances in the field of communication. the effective definition power of the network antennas, and the AAIC technique can be used in any system with network antennas in which interference is a potential problem.

  The described AAIC techniques can be implemented in any number of modules.

  Each module can be implemented in any of a number of ways, its implementation not being limited to the execution of the workflow as described above with precision. The AAIC processes described above and represented in flow charts and diagrams can be modified in any way to fulfill the functions described here.

  Of course, the various functions of the AAIC method and apparatus can be distributed in any way between any number (for example one or more) of modules or hardware units and / or software, systems or systems. computer circuits or processing.

  The AAIC processing module (s) may be integrated into a stand-alone antenna / receiver system or may operate separately and be connected to any number of devices, workstation computers, server computers, or storage devices. data via any communications medium (eg network, modem, direct connection, etc.). The AAIC system can be implemented by any number of devices and / or computer systems (personal or other computers) or processing. The computer system can include any commercially available operating system, any software (eg, communication software, among others) commercially available and / or customized and any type of devices. input (for example, radio receiver, among others).

  It is obvious that the software of the AAIC system can be implemented in any desired computer language and can be developed by the specialist in computer techniques and / or programming from the functional description given here and proposed flowcharts on the drawings. In addition, the AAIC software can be made available or distributed via any suitable medium (eg stored on devices, such as CD-ROM and diskette, downloaded from the Internet or from another network (eg via packets and / or carrier signals), downloaded from a telematic service (eg via carrier signals), or other conventional broadcast mechanisms .

  The AAIC output may be presented to the user and / or other processing modules in any manner using a digital and / or visual and / or audio presentation, and / or electronic data. In addition, all references herein to software performing different functions generally refer to computer systems or processors performing these functions under software control. The computer system may alternatively be implemented in hardware form or using other processing circuits. The different functions of the AAIC process can be distributed in any way between any number (for example, one or more) of modules or hardware units and / or software, computer or processing systems or circuits, computer or processing systems that can be located locally or remotely from one another and communicate through any suitable communications medium (eg local area network (LAN), long distance network (WAN), network internal, Internet, cable connection, modem link, wireless,). The software and / or processes previously described and represented in the flowcharts and diagrams can be modified in any way to fulfill the functions described herein.

  The system of the present invention may be implemented using any of a variety of hardware and software configurations and is not limited to any particular configuration. For example, the AAIC processing may be performed by means of a signal received via a suitable network antenna and any of the necessary antenna elements needs of a particular system, beam, angle scanning and antenna gain.

  Virtually all applications using directional antennas, such as phased array antennas, can benefit from the use of the described AAIC technique. These applications include, but are not limited to, communications, navigation, and radar systems, such as future generations of Global Positioning Systems (GPS), GPS augmentation systems, wireless telephony, systems satellite communications, the global multi-mission services platform (GMSP), systems using code division multiple access multiplexing (CDMA) and other communications systems.

  Although the foregoing description has focused on preferred embodiments of new and improved methods and apparatuses for adaptively performing algebraic deletion of pests, the present invention is of course not limited to the particular examples described and illustrated herein, and those skilled in the art will readily understand that it is possible to make many variations and modifications without departing from the scope of the invention. Similarly, although specific terms are used here, they are used in a purely generic and descriptive sense, and in no way limiting. may be

  configured any size; it particular number of elements using is not limited to receiving antenna one and any number provision adapted to meet such as width of

Claims (10)

  A method of suppressing interference of an interesting signal received by a phased array antenna comprising multiple antenna elements, characterized in that it comprises the steps of: (a) generating respectively a plurality of signal signals; antenna elements that reflect the reception of the interesting signal and noise at the multiple antenna elements; (b) adaptively determining the direction of one or more sources of noise received at the same time as the signal of interest; (c) constructing, for each source of noise, a directional noise vector corresponding to the direction of the noise source; (d) calculating a projection operator that transforms each directional noise vector into a null vector; (e) applying the projection operator to the antenna element signals to eliminate the parasites therefrom; and (f) restoring the signal of interest by compensating individually in the antenna element signals for an interesting signal distortion due to the application of the projection operator.   2. Method according to claim 1, characterized in that: the phased array antenna comprises M antenna elements; the projection operator is a matrix of MxM A of rank M-k, where k is the number of parasite sources identified during step (b) and k <M; and the matrix A is calculated to satisfy the expression Ain = 0 for n = 1 to k, where in represents the directional noise vectors.   3. Method according to claim 1, characterized in that step (b) comprises determining weighting values corresponding to the directions of the sources of parasites.   4. Method according to claim 3, characterized in that the weighting values are determined without execution of a beam forming operation using the weighting values.   Apparatus for suppressing noise from an interesting signal, characterized in that it comprises: a phased array antenna (202) comprising multiple antenna elements which respectively generate a plurality of antenna element signals reflecting the reception of the interesting signal and parasites at the antenna elements; and a processor (206) configured to adaptively determine the direction of one or more sources of noise received at the same time as the signal of interest; to construct, for each source of noise, a directional noise vector corresponding to the direction of the source of noise; to calculate a projection operator that transforms each directional noise vector into a null vector; for applying the projection operator to the antenna element signals to eliminate interference from the antenna element signals; and to restore the interesting signal by compensating individually in the antenna element signals for an interesting signal distortion due to the application of the projection operator.   An apparatus according to claim 5, characterized in that: the phased array antenna (202) comprises M antenna elements; and the processor (206) calculates the projection operator as a matrix of MxM A of rank Mk, where k is the number of noise sources and k <M, the matrix A being calculated to satisfy the expression Ain. = 0 for n = 1 to k, where in represents the directional noise vectors.   An apparatus according to claim 5, characterized in that the processor (206) adaptively determines weighting values corresponding to the directions of the noise sources.   An apparatus according to claim 7, characterized in that the processor (206) determines the weighting values without performing a beamforming operation using the weighting values.   Apparatus for suppressing noise from a signal of interest, characterized in that it comprises: a phased array antenna (202) comprising multiple antenna elements which respectively generate a plurality of antenna element signals reflecting the reception of the interesting signal and parasites at the antenna elements; means for adaptively determining the direction of one or more sources of noise received at the same time as the signal of interest; means for constructing a directional noise vector corresponding to the direction of each source of noise; means for calculating a projection operator that transforms each directional noise vector into a null vector; and means for applying the projection operator to the antenna element signals to eliminate interference from the antenna element signals; and means for restoring the signal of interest by individually compensating in the antenna signals a distortion of the signal at the application of the operator of claim 9, characterized in antenna phased array (202); and the means for calculating the projection operator computes the projection operator in the form of an MxM matrix A of rank Mk, where k is the number of noise sources and k <M, the matrix A being calculated to satisfy the expression Ai ,, = O for n = 1 to k, in which i represents the directional noise vectors.   Apparatus according to claim 9, characterized in that the means for adaptively determining the direction of the noise sources adaptively determines weighting values corresponding to the directions of the noise sources.   Apparatus according to claim 11, characterized in that the means for adaptively determining the direction of the noise sources determines the weighting values without performing a beamforming operation using the weighting values.   interesting projection.   10. Apparatus what: the antenna comprises M elements due