RU2629921C1 - Method of adaptive signal processing in modular phase antenna lattice - Google Patents

Abstract

FIELD: radio engineering, communication.

SUBSTANCE: receive signals from a known direction by an even number of modules arranged symmetrically with respect to the phase center of the modular phased array antenna with an amplitude and complex conjugate phase distribution that is symmetric with respect to the phase center of the opening. For each pair of symmetrically located modules, the sum and difference signals of the pairs of modules are formed. Summarize the total signals of the pairs of modules by the power, forming the initial sum signal of the modular phased array antenna. The covariance matrix of the difference signals of the module pairs and the covariance coefficient vector of the initial sum signal of the modular phased array antenna and the difference signals of the module pairs are found. Summarize the covariance matrix of the difference signals of pairs of modules with a diagonal matrix. The greater the weight of the diagonal, the smaller the values of the correcting phases. A matrix of coefficients is generated and the phase vector of the pairs of modules is determined by multiplying the inverse coefficient matrix by the covariance vector of the initial sum signal of the modular phased array antenna and the difference signals of the module pairs. Change, according to the detected phase of the correcting phases, the phases of the signals of the modules and sum the signals of the pairs of modules with the changed phases, forming the output signal of the modular phased array antenna.

EFFECT: possibility of adaptive signal processing in a modular phased array that realizes signal processing based on real arithmetic, while maintaining its speed.

5 dwg

Description

The invention relates to the field of radio engineering and communication and can be used in radio communication systems operating in a complex jamming environment.

Adaptive antenna arrays are known that implement an algorithm for maximizing the output ratio of the useful signal power to the sum of the interference and noise powers [1. Monzingo R.A., Miller T.U. Adaptive antenna arrays. Introduction to the theory. M .: Radio and communication, 1986, p. 80-86, 179-240]. For the operation of an adaptive antenna array of this type, a priori information about the direction of arrival of the useful signal is used. During the operation of adaptive antenna arrays of this type, a complex covariance matrix of signal and interference is formed, after which the original vector of complex weight coefficients of the adaptive antenna array is converted to a vector of complex weight coefficients, ensuring the maximum signal / (noise + noise) ratio is reached.

When implementing this method for the formation of a complex covariance matrix of the signal and interference, it is necessary in each channel of the adaptive antenna array to select two quadrature components of the signal and to carry out cross-covariance processing of the signals received by the channels of the adaptive antenna array. This complicates the system; in addition, with an increase in the number of channels, the circulation time of the covariance complex matrix increases.

To overcome these shortcomings, the number of adaptive channels is reduced by combining antenna channels into receiving modules [2. Pat. 2491685 (RU). A system for simplifying the processing of a reconfigurable beam-forming circuit in a phased array for a telecommunication satellite / Craig ED, Stirland SD H01Q 1/28. Publ. 08/27/2008. Bull. No. 24]. In the extreme case, the adaptive grating degenerates into a two-channel system of adaptive interference compensation, which contains only one correlator and is able to deal with one interference [3. Protection against radio interference / M.I. Maximov, M.P. Bobnev, B.Kh. Krivitsky [et al.]; under the editorship of M.I. Maksimova. - M .: Owls. Radio, 1976. - 496 p.].

In addition to the above disadvantages, the implementation of signal processing in adaptive antenna arrays requires controlling the amplitudes and phases of the signals in the channels of adaptive antenna arrays.

The closest in technical essence to the claimed method is a method for implementing an adaptive antenna array in arithmetic of real numbers, described in [4. Dzhigan V.I. Computationally effective linearly bounded complex RLS-algorithm in the arithmetic of real numbers // Reports of the 14th International Conference “Digital Signal Processing and its Applications (DSPA-2012)” (Moscow, March 28-30, 2012). Moscow. 2012. Volume 1]. In accordance with this method, adaptive channels are arranged symmetrically in the array, and the initial amplitude-phase distribution in them is complex conjugate relative to the phase center of the antenna. This allows you to reduce the time of transients in the adaptive antenna array by 1.5 ... 2 times due to the implementation of the adaptation algorithm in the operations of real arithmetic.

The method consists in the fact that the received every M-th channel module adaptive antenna array with a symmetric arrangement of antenna channels with respect to phase center signals for a predetermined position of the maximum of the radiation pattern representing a mixture of the desired signal, interference and noise, are summed in M _and modules to obtain output signals of M _a modules. The output signals of the modules are divided by power into the transmitted and branch parts. The signals corresponding to the branched part of the power used to generate the covariance matrix of interference signals and C and determines complex weighting coefficients as a vector string J of the M elements _and the formula J = J ⁰ (C + αI) ^-1, where I - the identity matrix , J ⁰ is the initial vector, which is complex conjugate relative to the phase center of the lattice, with which the signals of the modules are summed. The signals corresponding to the transmitted part of the power are summed using the found complex weight coefficients J, forming the output signal of the adaptive antenna array.

For adaptive signal processing in this method, it is required to control both the amplitudes and the phases of the complex weight coefficients, which in practice limits the application of this method only to digital antenna arrays.

The problem to which the invention is directed is the possibility of adaptive signal processing in a modular phased antenna array that implements signal processing based on actual arithmetic, while maintaining its speed.

To solve this problem, a method for adaptively processing signals in a modular phased antenna array is proposed, consisting in the fact that the reception of signals from a known direction is carried out by an even number of modules located symmetrically relative to the phase center of the modular phased antenna array with an amplitude amplitude and complex conjugate phase opening symmetrical with respect to the phase center distribution. In contrast to the prototype method, for each pair of symmetrically located modules form the total and difference signals of the pairs of modules. The total signals of the pairs of modules are summed over the power, forming the initial total signal of the modular phased antenna array. Find the covariance matrix of the difference signals of the pairs of modules and the vector of the covariance coefficients of the initial total signal of the modular phased antenna array and the difference signals of the pairs of modules. The covariance matrix of the difference signals of pairs of modules with a diagonal matrix is summarized, the more the weight of the diagonal, the lower the values of the correcting phases. A matrix of coefficients is formed and the vector of the correcting phases of the signals of the pairs of modules is determined by multiplying the inverse matrix of coefficients by the vector of the covariance coefficients of the initial total signal of the modular phased array antenna and the difference signals of the pairs of modules. According to the found vector of correcting phases, the phases of the signals of the modules are changed and the signals of the pairs of modules with the changed phases are summed, forming the output signal of the modular phased antenna array.

The technical result of the invention is adaptive signal processing in a modular phased array antenna that implements signal processing based on actual arithmetic, while maintaining its speed.

A comparative analysis of the claimed method and the prototype method shows that the claimed method differs in that the set of actions is changed, namely, five actions are introduced:

- for each pair of symmetrically located modules form the total and difference signals of pairs of modules;

- summarize the power of the total signals of the pairs of modules, forming the original total signal of a modular phased antenna array;

- find the covariance matrix of the difference signals of the pairs of modules and the vector of the covariance coefficients of the initial total signal of the modular phased antenna array and the difference signals of the pairs of modules;

- summarize the covariance matrix of the difference signals of pairs of modules with a diagonal matrix, and the larger the weight of the diagonal, the lower the value of the correcting phases;

- form a matrix of coefficients and determine the vector of the correcting phases of the signals of the pairs of modules by multiplying the inverse matrix of coefficients by the vector of the covariance coefficients of the initial total signal of the modular phased array antenna and difference signals of the pairs of modules;

and the mode of action associated with the formation of the output signal of the modular phased antenna array is changed:

- change, according to the found vector of the correcting phases, the phases of the signals of the modules and sum the signals of the pairs of modules with the changed phases, forming the output signal of the modular phased antenna array.

The introduction of five actions and changing the mode of one action allows, in comparison with the prototype method, to provide a technical result consisting in adaptive signal processing in a modular phased array antenna that implements signal processing based on actual arithmetic while maintaining its speed.

The combination of distinguishing features and properties of the proposed method for processing signals in a modular phased array antenna are unknown from the literature, and information sources containing information about similar technical solutions having features similar to those distinguishing the claimed solution from the prototype, as well as properties that match properties of the proposed solution, therefore, we can assume that it has significant differences, follows from them in an unobvious way and, therefore, corresponds to patentability criteria “novelty” and “inventive step”.

The essence of the proposed method is disclosed by figures 1-5.

The figure 1 shows the structural diagram of a modular phased antenna array that implements the proposed method.

Figures 2-4 show the radiation patterns of a modular phased array antenna with various sets of phase distributions.

Figure 5 presents two phase distributions corresponding to the radiation patterns shown in figure 2 and figure 3.

When implementing the proposed method for adaptive signal processing in a modular phased array, the following sequence of operations is performed:

- receiving signals from a known direction is carried out by an even number of modules located symmetrically with respect to the phase center of the modular phased array antenna with symmetric with respect to the phase center aperture amplitude and complex conjugate phase distribution -1;

- for each pair of symmetrically located modules form the total and difference signals of pairs of modules -2;

- summarize the power of the total signals of the pairs of modules, forming the original total signal of the modular phased antenna array -3;

- find the covariance matrix of the difference signals of the pairs of modules and the vector of covariance coefficients of the initial total signal of the modular phased antenna array and the difference signals of pairs of modules -4;

- summarize the covariance matrix of the difference signals of pairs of modules with a diagonal matrix, and the larger the weight of the diagonal, the lower the value of the correcting phases -5;

- form a matrix of coefficients and determine the vector of the correcting phases of the signals of the pairs of modules by multiplying the inverse matrix of coefficients by the vector of the covariance coefficients of the initial total signal of the modular phased array antenna and the differential signals of the pairs of modules -6;

- change, according to the found vector of the correcting phases, the phases of the signals of the modules and sum the signals of the pairs of modules with the changed phases, forming the output signal of the modular phased antenna array - 7.

The structure of a modular phased antenna array (PAR) (Figure 1) includes 2M antenna elements (AE) 1, divided into groups of N elements each, connected to 2P receiving modules (sublattices) (PM) 2. Each PM 2 contains a receiver adder the module and N phase shifters connected by inputs to the N antenna elements, and the outputs to the adder of the module (phase shifters and the adder of the receiving module in figure 1 are not indicated). The outputs symmetric with respect to the phase center of the modular headlight PM 2 are connected through phase shifters (PV) 3 to the inputs of the corresponding total-difference converters (SRP) 4. The total outputs of the SRP 4 are connected to the inputs of the adder of the modular headlight (Σ) 5. At the output of the adder, the modular headlight Σ 5 a directional coupler (BUT) is installed 6. The differential outputs of the PSA 4 are connected to the inputs of the correcting phase calculator (VKF) 7. The outputs of the VKF 7 are connected to the corresponding control inputs of the PV 3. Control inputs PM 2 (control inputs of the phase shifters, set data in the channel of each AE 1) are connected to the corresponding outputs of the phase calculator (WF) 8. The first output of HO 6 is the output of the modular phased array and is designated as Output. 9. The second output of HO 6 is connected to the input of the VKF 7.

Consider the operation of a modular phased array in which adaptive signal processing is performed.

In accordance with the information about the direction of the main maximum of the radiation pattern (LH) of the modular headlamp coming from the outputs of the WF 8 phase calculator to the control inputs of PM 2 (to the control inputs of the phase shifters of the receiving modules), the main maximum of the LH is oriented in the direction of the useful signal. For each position of the maximum of the ND, the signals received by the AE 1 antenna elements, which are a mixture of the useful signal, interfering signals, and the intrinsic (thermal) noise of the headlamp channels, are fed to the PM 2 inputs. The signals from the PM 2 outputs symmetrically located relative to the phase center of the modular phased array are fed to the inputs of the phase shifters ФВ 3. The signals from the outputs of the phase shifters ФВ 3 are fed to the inputs of the corresponding total-difference converters СРП 4, at the outputs of which the total and difference signals of the PM 2 pairs are formed. The output signals from the outputs of the sum-difference converters СРП 4 are fed to the inputs of the adder of the modular ФА Σ 5, the output of which is connected to the directional coupler НО 6. The first output of the directional coupler НО 6 is the output of the modular ФАР Output. 9. Therefore, at its output, the initial total signal of the modular headlamp is formed. The differential signals of the PM 2 pairs from the outputs of the SRP 4 total-difference converters and the branched initial total signal of the modular PAR from the second output of the NO 6 directional coupler are fed to the corresponding inputs of the correcting phase calculator VKF 7, in which the covariance matrix of the difference signals of the pairs of receiving PM 2 and covariance vector of the initial total signal of the modular PAR and differential signals of pairs of receiving modules PM 2. In the calculator of correcting phases VKF 7, the covariance matrix of the difference signal in par receiving modules PM 2 is added to a diagonal matrix (the larger the weight of the diagonal, the smaller value correcting phases). A matrix of coefficients is also generated in the calculator of corrective phases VKF 7 and a vector of corrective phases (corrective phases) of the signals of the pairs of receiving modules PM 2 is determined by multiplying the inverse matrix of coefficients by the vector of covariance coefficients of the initial total signal of the modular PAR and differential signals of pairs of receiving modules PM 2. Vector of corrective the phases of the signals of the pairs of receiving modules PM 2 in the calculator of the correcting phases VKF 7 is converted into the vector of the correcting phases of the signals of the receiving modules PM 2 by shifting by 18 0 ° corrective phases for symmetric relative to the phase center of the aperture PM 2. According to the found vector of correcting phases of the signals of the receiving modules PM 2, phase shifters PV 3 change the phase of the signals. The PM 2 signals with the changed phases, passing the total-difference converters of the SRP 4, are summed in the adder of the modular phased array Σ 5 and fed into the first channel of the directional coupler HO 6, at the output of which the output signal of the modular phased arrester is generated. 9.

To justify the process of adaptive signal processing in a modular phased array, we state the following.

.Let there be 2M ≥ 2PN - elementary modular PAR (N is the number of antenna elements in one receiving module; 2P is the number of receiving modules). We will also assume that the PAR architecture is symmetric, and for signal processing, the amplitude distribution symmetrical with respect to the phase center is used, which is described by the row vector

.

Suppose also that the signal is received by the modular headlamp from a known direction θ ₀ , and the sources of interference from a priori unknown directions θ _i (i = 1, 2, ..., I).

The task is to increase the signal / (interference + noise) (SINR) ratio by selecting the phase values of the PAR modules.

The pattern of the lattice under consideration is described by an expression of the form

where ƒ ₀ (θ) is the DD of a single element of the PAR;

ψ _p is the desired phase of the rth receiving module;

And _{p, n} is the actual amplitude of the nth element in the composition of the rth receiving module;

x _{p, n} is the coordinate of the phase center of the nth element in the composition of the rth receiving module.

If we take into account the symmetry of the placement of the receiving modules in the grating and the symmetry of the amplitude distribution, and also assume that the desired phases of the symmetric receiving modules are opposite in phase, then expression (1) can be converted to:

Suppose that the desired phases are sufficiently small. Then in expression (2) we can assume that cosψ _p ≈ 1, a sinψ _p ≈ψ _p . (Such an assumption is necessary for the transition to real arithmetic, but does not mean a rejection of the phase control of the modular headlamp.) In this case, we obtain:

In expression (3), the FF ₀ (θ) is introduced, which corresponds to the modular HEADLAM at zero phases of receiving modules.

Taking into account the introduced notation, the signal / (noise + noise) ratio in the considered modular headlamp can be represented as

where T (θ) describes the spatial distribution of interference and thermal noise:

where δ ² is the dispersion of thermal noise;

P _i is the interference power.

It follows from expression (4) that the formulated problem consists in minimizing the denominator in expression (4) for the SINR while restricting the values of the phases of the modules.

Constraints on phase values can be represented as:

It follows that the formulated problem reduces to determining the phases ψ _p that provide the minimum value of the Lagrange functional:

where α is the Lagrange multiplier.

The components of the functional gradient (7) can be represented as:

Denote

where m, p = 1, 2, ..., R.

Then the optimal sets of phases can be found from the solution of the system of linear algebraic coefficients (SLAE):

It is convenient to represent the system of equations (11) in matrix form:

where I is the identity matrix;

ψ is a row vector of the desired phases for half of the PAR modules.

It should be noted that the elements of the matrix S and the vector B are real; the dimension of the matrix at the desired phases and the vector of the right parts corresponds to half of the total number of receiving modules. This provides a high speed of solving the system of equations (12) in comparison with similar systems using complex arithmetic without restrictions on complex weighting coefficients.

For α = 0, for the system of equations (12), there are no restrictions on the phase values, however, the conditions for the smallness of the phases are violated, which may lead to the fact that the used PAR model (expression (3)) will be inadequate. On the contrary, as α → ∞, the contribution of the matrix S to the resulting solution will be insignificant. Therefore, the parameter α must be selected in such a way as to provide a compromise between restrictions on the acceptable values of the phases of the receiving modules and the sensitivity of the system to an interference environment.

Let us now consider how SLAE can be formed (11) by acting on signals.

At the output of each antenna element AE 1 of the modular headlight, a superposition of the signal, interference, and thermal noise is formed:

where y (t), s _i (t) and n _{p, n} (t) are the envelopes of the signal, noise, and thermal noise in the nth channel of the rth receiving module PM 2 (n = 1, 2, ..., N; p = 1, 2, ..., P).

When subtracting the signals received by symmetrically spaced channels (PM 2), we get:

Thus, the signal component is excluded from expression (14).

We perform the summation of the received difference signals in each of the receiving modules PM 2:

where n ' _p (t) is the envelope of thermal noise at the output of the rth receiving module PM 2.

Find the covariance coefficients of the difference signals of the receiving modules PM 2:

In expression (16), the “bar above” denotes the statistical averaging of received signals over a selected period of time.

An analysis of expression (16) allows us to conclude that in most practical cases it can be significantly simplified by discarding non-essential terms. So, the thermal noise n ' _p (t) at the output of the receiving modules PM 2 are uncorrelated. The same statement is true with respect to covariances of noise and thermal noise. Therefore, expression (16) can be reduced to the form:

имеет ярко выраженную диагональ, которая пропорциональна мощностям соответствующих помех, т.е.:Obviously, in most practically important cases, the sources of interference are at different distances from the antenna, can move at different angular speeds, etc. Therefore the matrix

has a pronounced diagonal, which is proportional to the power of the corresponding interference, i.e.:

Given expression (5) and the additive properties of noise, it can be argued that

From the theory of antenna arrays it is known that this matrix is positive definite. This means that a solution to the system of equations (11) must exist.

The vector of the right-hand sides of the system of equations (11) is a time-averaged product of the difference signal of the PM 2 receiving module Δ _p (t) and the total signal of the original modular PAR, provided that the signal component is excluded.

The output signal of the original modular headlamp contains the components of the signal, interference and thermal (internal) noise:

where n ' ₀ (t) is the envelope of noise at the output of the modular headlamp.

A statistically averaged product of signals (20) and (15) can be represented as:

Taking into account the comments made on the averaging of statistically unrelated signals, we write expression (21) in a simplified form:

Using the notation introduced earlier, we can verify that

The second term in expression (23) vanishes if there is no correlation between the signal and the noise. This assumption is common in the theory of adaptive antenna arrays. Therefore, when this condition is met, we obtain:

Thus, the system of equations (12) can be obtained by averaging the product of the signals of the modular headlamp and the difference signal obtained by subtracting the signals of the symmetrically located receiving modules PM 2.

The obtained relations confirm that the system of equations (12) has a solution and can be formed as a result of the correlation processing of signals of the modular PAR.

To confirm the operability and effectiveness of the patented method of adaptive signal processing in a modular phased array, computer simulation was carried out. As an example, consider 2M = 32, an elementary linear antenna array (AR) with a spacing between elements of d = 0.5λ.

The bottom of the antenna element is described by a complex function

The interference environment is set using two sources of interference located in the directions θ _1.2 = (8.8 °; -12.9 °). The signal (useful signal) comes from the direction θ ₀ = 0.

As the amplitude distribution in the aperture, the falling distribution “cosine on the pedestal 0.7” was chosen.

In figures 2-4 presents the AR AR obtained by solving the system of equations (12) with α = 0.2, 2 and 0, respectively. The dashed line corresponds to the AR of the AR without correction of the phase distribution; the solid curve indicates the AR of AR after adaptation. From a comparison of the graphs in figures 2-4, it follows that restrictions on the phase values are necessary, since the DN in figure 4 is "destroyed" due to the out-of-phase elements. For small values of the parameter α, the depth of noise suppression is the best, but at the same time, there is an increase in individual side lobes by about 6 dB (figure 2), which is consistent with the results of [5. Vendik O.G., Parnes M.D. Antennas with electronic beam movement. Introduction to the theory. M.: Saynes-press, 2002. - 302 p.] For adaptive antenna arrays with phase control. An increase in the parameter α leads to the formation of less deep zeros of the MD, however, bursts of the side lobes also decrease to about 4 dB (figure 3).

Figure 5 shows two phase distributions corresponding to the NDs shown in Figure 2 and Figure 3. These results show that an increase in the parameter α leads to a decrease in the phase spread in the receiving modules.

A modular phased antenna array that implements the patented method can be built on the basis of microwave devices widely used in development and well-developed in the production of microwave devices (Fig. 1, elements 1-7): antenna elements controlled by analog or digital phase shifters, sum-difference converters, directional couplers and adders [see, for example: 6. Active phased array antennas / Ed. DI. Voskresensky and A.I. Kanaschenkova. M .: Radio engineering, 2004. S. 66-82, 121-130; 7. Microwave devices and antennas / Ed. DI. Voskresensky. Ed. 2nd, add. and reslave. M .: Radio engineering. 2006. S. 87-96]. To create electronic storage, computing and control units (elements 8, 9 in Fig. 1), there is a developed elemental base, in particular programmable logic integrated circuits and digital signal processors, which provide the implementation of control and data processing functions. The domestic signal controller 1892 VMZT has such capabilities [8. Pletneva I.D. Implementation of control algorithms for adaptive antenna arrays based on a digital signal controller // Izv. universities. Electronics, 2009, No. 3, p. 61-67].

The implementation of the proposed method is much simplified in relation to digital antenna arrays.

The above materials confirm compliance with the criterion of "industrial applicability" of the proposed method.

Thus, the patented method of adaptive signal processing in a modular phased antenna array is practically feasible and provides the declared technical result, which consists in adaptive signal processing in a modular phased antenna array that implements signal processing based on actual arithmetic, while maintaining its speed.

Claims (1)

The method of adaptive signal processing in a modular phased antenna array, which consists in the fact that the reception of signals from a known direction is carried out by an even number of modules located symmetrically relative to the phase center of the modular phased antenna array with an amplitude amplitude and complex conjugate phase distribution symmetrical with respect to the phase center, characterized in that for each pair of symmetrically located modules form the total and difference signals of pairs of modules, summarize the power the sum signals of the pairs of modules, forming the initial sum signal of the modular phased antenna array, find the covariance matrix of the difference signals of the pairs of modules and the vector of covariance coefficients of the initial sum signal of the modular phased antenna array and the difference signals of the pairs of modules, summarize the covariance matrix of the difference signals of the pairs of modules with a diagonal matrix, and the larger the weight of the diagonal, the lower the values of the correcting phases, form a matrix of coefficients and determine the vector their phases of the signals of the pairs of modules by multiplying the inverse matrix of coefficients by the vector of covariance coefficients of the initial total signal of the modular phased array antenna and the difference signals of the pairs of modules, change, according to the found vector of the correcting phases, the phases of the signals of the modules and sum the signals of the pairs of modules with the changed phases, forming the output signal modular phased array antenna.

Images