RU2803416C1 - Method for forming weight coefficients in the processing channels of a broadband adaptive antenna array - Google Patents

Abstract

FIELD: antenna technology.

SUBSTANCE: proposed method for the formation of weight coefficients in the processing channels of a broadband adaptive antenna array consists in the fact that the additive mixture of a broadband signal and interference using bandpass filters is divided into narrowband frequency intervals, for each of which at certain frequencies ω_k the optimal vector of weight coefficients W (ω_k) is found, for other frequencies of processing the spectrum of the useful broadband signal, quasi-optimal weighting coefficients are found based on various interpolation functions. Moreover, the optimal vector of weight coefficients is found with a smaller frequency interval at those frequencies of the useful broadband signal where the spectral density of the useful broadband signal is low and with a large interval at those frequencies where the spectral density of the useful broadband signal is high.

EFFECT: rational distribution of frequencies in which the vector of weight coefficients is optimally determined, taking into account the spectral density of the useful broadband signal.

1 cl, 3 dwg

Description

The invention relates to antenna and computer technology and can be used in broadband radio communication systems, radar and radio navigation when receiving broadband signals (BPS) under conditions of exposure to broadband and narrowband interference.

There is a known method for forming a vector of weighting coefficients in an adaptive antenna array [1, p. 56, 2], containing N antenna elements. A device with quadrature channels is introduced into the channel of each antenna element, with the help of which the signal is divided into in-phase and quadrature components, and each of the components is subjected to the operation of multiplication by a weighting coefficient. The signals obtained after such processing are added in the adder. The values of the weighting coefficients are controlled using a signal processor.

However, in the case when the useful and interference signals are characterized by a frequency spectrum of significant width, a complex weighting coefficient must be selected for each of the frequencies, otherwise effective suppression of all interference components will not be ensured.

An adaptive antenna array is known [2]. Using filters, power measurement units, a comparison unit and a control unit, the output power of the common adder is minimized and maximized in the modes of noise suppression and useful signal selection.

The disadvantage of this adaptive antenna array is the inability to process NPS. Also, the approach to forming a vector of weighting coefficients is not considered.

To eliminate the shortcomings of devices that implement the classical method of spatial filtering of narrowband signals, a transversal filter or a multi-tap delay line is introduced, which provides interference suppression in the frequency band [1, p. 57-60].

However, the use of frequency-dependent weighting using a multi-tap delay line is associated with the selection and implementation of the necessary amplitude and phase characteristics of complex weighting coefficients. In the proposed analogues, devices that ensure the selection and implementation of the required amplitude-phase characteristics of complex weight coefficients are not considered.

The closest in technical essence to the proposed method is the method of forming weight coefficients for processing ShPS, described in [3]. This method was chosen as a prototype.

The essence of the prototype method is that the additive mixture of NPS and interference is divided into narrow-band frequency intervals using bandpass filters. For each frequency interval at a certain frequency ω _k , finding the optimal vector of weighting coefficients W(ω _k ) according to the criterion of the maximum signal/(interference + noise) ratio (SINR). For other processing frequencies of the spectrum of the useful NPS, quasi-optimal weighting coefficients are found based on various interpolation functions, in particular based on piecewise constant and piecewise linear interpolation. This method allows for spatial processing of NPS against the background of various noises.

The disadvantage of this method of forming weight coefficients is the impossibility of taking into account the spectral structure of the NPS, since the method does not describe the procedure for selecting frequencies in which the vector of weight coefficients is optimally determined.

The technical result of the proposed invention is aimed at expanding the functionality consisting in the rational distribution of frequencies in which the vector of weighting coefficients is optimally determined, taking into account the spectral density of the useful broadband.

To achieve the specified technical result in the method of forming weighting coefficients in the processing channels of a broadband adaptive antenna array, which includes separating an additive mixture of a broadband signal and interference using bandpass filters, they are divided into narrowband frequency intervals. For each of the frequency intervals at certain frequencies ω _k , finding the optimal vector of weighting coefficients W(ω _k ) according to the maximum SINR criterion. For other frequencies of processing the spectrum of a useful broadband signal, quasi-optimal weighting coefficients are found based on various interpolation functions. In this case, the optimal vector of weighting coefficients is found with a smaller frequency interval at those frequencies of the useful broadband signal at which the spectral density of the useful broadband signal is low and with a larger interval at those frequencies at which the spectral density of the useful broadband signal is high.

In fig. Figure 1 shows the spectrum of a broadband chirp signal with the designation of its informative part.

In fig. Figure 2 shows the spectrum of a broadband chirp signal indicating the frequencies in which the vector of weighting coefficients is optimally determined based on the approach proposed in the prototype method.

In fig. Figure 3 shows the spectrum of a broadband chirp signal indicating the frequencies at which the vector of weighting coefficients is optimally determined based on the proposed method.

The method is implemented as follows.

1. When choosing the type of useful NPS, its spectrum is calculated based on the Fourier transform [4]:

where S(t) is the type of selected ShPS;

ω=2πƒ - circular frequency, [rad/s];

ƒ - linear frequency [Hz].

2. Next, the ratio of the power of the useful NPS and the interfering effect at the output of the matched filter is determined under the assumption that the signal spectrum |С(ω)| is rectangular and equal to the constant C within the signal band W and zero outside it, and the amplitude transfer function of the filter is uniform within the signal band W and equal to zero outside it. It is also allowed that its non-zero value is equal to one. Consequently, the interference, interpreted as a random process, passes to the output of the filter without changing its power J, while the filtered value of the noise power will be N ₀ W. On the other hand, the filter is consistent with the signal and, therefore, coherently sums all the harmonic components of the signal, providing maximum output amplitude value:

where the uniformity of the spectrum in the signal band W is taken into account, and the doubling is due to the contribution of “negative” frequencies. Using similar notation, the signal energy is calculated using Parseval's theorem as:

Thus, the SINR at the output of the matched filter is:

From relation (4) it is clear that the lower the energy E of the useful NPS, the lower the SINR at the output of the matched filter.

In this way, the SINR is calculated for each frequency of the useful broadband with a certain discrete value, the value of which depends on the required SINR at the output of the antenna array, and the low energy level of the useful broadband is detected.

3. Since the signal spectrum is infinite, there is a certain frequency range, called frequency deviation, in which the entire informative signal is concentrated (Fig. 1). Outside this range, it makes no sense to optimally determine the vector of weighting coefficients, since this does not play a special role in restoring the spectrum of the useful NPS. In fig. 2 it can be seen that in the prototype method the optimal vector of weighting coefficients is determined at equal intervals of the spectrum of the useful NPS and the structure of the spectrum itself is not taken into account.

4. Based on the calculation results obtained from relation (4), frequencies are rationally determined in which the optimal vector of weighting coefficients must be determined at a more frequent interval (the spectral density of the useful NPS is low), and in which at a more rare interval (the spectral density of the useful NPS is high ). In this case, the total number of processing channels does not change (Fig. 3).

5. Next, for other frequencies of the useful broadband spectrum, the vector of weighting coefficients is interpolated by various functions.

Thus, the method of forming weight coefficients in the processing channels of a broadband adaptive antenna array makes it possible to ensure the required level of SINR at the output of the antenna array due to the rational distribution of frequencies in which the vector of weight coefficients is optimally determined and thereby achieve a technical result consisting in expanding functionality, namely in the rational distribution of frequencies in which the vector of weighting coefficients is optimally determined, taking into account the spectral density of the useful broadband.

Literature

1. Monzingo R.A., Miller T.W. Adaptive antenna arrays: Introduction to theory: Transl. from English - M.: Radio and communication, 1986. - 448 p.

2. AC 1548820, 1990

3. RU 2466482, 2012

4. Ipatov V. Broadband systems and code separation of signals. Principles and applications, 2007. - 488 p.

Claims (1)

A method for forming weighting coefficients in the processing channels of a broadband adaptive antenna array, which consists in dividing an additive mixture of a broadband signal and interference using bandpass filters into narrow-band frequency intervals, for each of which, at certain frequencies ω _k , finding the optimal signal/to-ratio maximum criterion. (interference + noise) vector of weighting coefficients W(ω _k ), for other frequencies of processing the spectrum of a useful broadband signal, quasi-optimal weighting coefficients are found based on various interpolation functions, characterized in that the optimal vector of weighting coefficients is found with a smaller frequency interval at those frequencies of the useful broadband signals at which the spectral density of the useful broadband signal is low and with a large interval at those frequencies at which the spectral density of the useful broadband signal is high.

Images