RU2429494C1 - Detection method of multiple narrow-band radio signals in broad band of frequencies - Google Patents

Abstract

FIELD: radio engineering. ^ SUBSTANCE: essence of detection method of multiple narrow-band radio signals in broad band of frequencies consists in performance of analogue-to-digital conversion of broad-band input process (BIP), accumulation of complex BIP readouts and in-series formation of samplings with BIP readouts, calculation of amplitude spectrum of formed sampling with BIP readouts, calculation of detection threshold value for each component part of amplitude spectrum as sum of smoothed estimate of spectrum in weighing window sliding along frequency axis and function from their mean-square deviation, comparison of amplitudes of spectrum component parts with detection threshold value, storage of comparison results of amplitudes of spectrum components with detection threshold value, taking the decision on detection of radio signals in investigated frequency band on the basis of majority processing of stored comparison results of amplitudes of spectrum components with detection threshold value, which were obtained at analysis of six in-series taken samples with BIP readouts, and output of detection results. ^ EFFECT: reducing the influence of signal-interference environment. ^ 7 dwg

Description

The invention relates to the field of radio engineering and can be used in radio communications to detect short-term radio signals in the frequency band exceeding the width of their spectrum by four times or more.

A known method of separating discrete components in the spectrum of the signal [RF patent No. 2331893, IPC G01R 23/16. The method of separation of discrete components in the spectrum of the signal and the device for its implementation [Text] / Garin V.Yu., Ovsyannikov V.N., applicant and patent holder Federal State Unitary Enterprise “Central Scientific Research Institute named after Academician A.N. Krylov” - No. 2006138732/28, declared November 2, 2006, publ. 08/20/2008.], Including narrow-band spectral analysis of the signal, the formation of the threshold level of the spectrum, smoothing the spectrum, the formation of the detection threshold as the sum of the spectral components of the smoothed spectrum with a threshold level and the selection of spectral components that exceed the detection threshold.

The disadvantages of this method is the reduction in the probability of correct detection (R _software ) with increasing dynamic range of amplitudes of the discrete components of the spectrum in the studied frequency band. This drawback is due to the fact that the constant value of the threshold level used in the calculation of the detection threshold is the same for all components of the spectrum, which, if there are discrete components of a spectrum with a large amplitude in the studied frequency band, leads to the omission of discrete components of the spectrum with a small amplitude.

Closest to the proposed invention is a method of single-channel signal detection [Rembowski A.M., Ashikhmin A.V., Kozmin V.A. Radio monitoring: tasks, methods, tools. / Edited by A.M. Rembovsky. - M .: Hot line - Telecom, 2006 .-- 492 p.: Ill. - ISBN 5-93517-326-3 (p. 100-129)], including analog-to-digital conversion of a broadband input process (ball screw), the accumulation of complex ball screw samples and sequential sampling with ball screw samples, the calculation of the amplitude spectrum of the sample with ball screw samples , averaging the spectra obtained by processing at least sixteen consecutively taken samples with ball screw samples, calculating the detection threshold, comparing the amplitudes of the spectral components with the detection threshold, deciding whether to detect signals in the studied frequency band and the output of the detection results. This method is selected as a prototype.

Compared with the previous analogue in the prototype method, the dependence of P _software on the dynamic range of amplitudes of the radio signals present in the studied frequency band is reduced.

The disadvantages of the prototype method are:

1. The decrease in the R _{PO of} radio signals with an increase in the load of the investigated frequency band by narrow-band radio signals by more than 30% [Rembovsky AM, Ashikhmin AV, Kozmin VA Radio monitoring: tasks, methods, tools / Edited by AMRembovsky. - M .: Hot line - Telecom, 2006 .-- 492 p.: Ill. - ISBN 5-93517-326-3 (p. 110)].

The reasons for the lack are as follows:

the detection threshold value is calculated on the basis of numerical characteristics calculated taking into account all components of the ballscrew spectrum in the studied frequency range and has a constant value for all components of the ballscrew amplitude spectrum;

the detection threshold value depends on the number of radio signals present in the ballscrew.

2. The dependence of R _software on the transmission coefficient of the path of the radio receiving device (RPU). This disadvantage is due to the fact that the minimum value of the smoothed amplitude spectrum of ballscrews is selected as the value of the detection threshold [Rembovsky AM, Ashikhmin A.V., Kozmin V.A. Radio monitoring: tasks, methods, tools. / Edited by AMRembovsky. - M .: Hot line - Telecom, 2006 .-- 492 p.: Ill. - ISBN 5-93517-326-3 (p. 110)].

3. Dependence of R _software on the law of distribution of atmospheric, cosmic interference and intrinsic noise of the Russian State Technical University in the studied frequency band (Dolukhanov MP. Radio wave propagation. State Publishing House of Literature on Communications and Radio. Moscow. 1960 (p. 378, Fig. 6.2, p. 382, fig. 6.5)). The reason for this drawback, as well as the previous one, lies in the choice of the minimum value of the smoothed amplitude spectrum of the ball screw as the detection threshold.

4. Skipping radio signals, the duration of which is less than the duration of sixteen consecutive samples with ballscrew samples at a ratio of radio signal power to noise power (SNR) from 3 to 5 dB in the frequency band occupied by the detected radio signal. The reason for this drawback is the use of averaging spectra to reduce the likelihood of false alarm (P _RT ). Moreover, P _LT equal to from 0.01 to 0.05 can be achieved by processing at least sixteen consecutively taken samples with ball screws [Rembovsky AM, Ashikhmin A.V., Kozmin V.A. Radio monitoring: tasks, methods, tools. / Edited by AM Rembowski. - M .: Hot line - Telecom, 2006 .-- 492 p.: Ill. - ISBN 5-93517-326-3 (p. 111, Fig. 3.4)].

5. High error in determining the time of detection of a radio signal at SNR from 3 to 5 dB in the frequency band occupied by the detected radio signal. The reason for this drawback, as well as the previous one, is that it is necessary to process at least sixteen consecutively taken samples with ball screw samples to reduce P _RT .

EFFECT: reduction of the influence of signal-noise situation (SVE), transmission coefficient of an RPU path, law of distribution of atmospheric, space interference and intrinsic noise of an RPU in the studied frequency band on the R _{PO of} narrow-band radio signals in the frequency band four times or more their spectral width as well as reducing the duration of detectable narrow-band radio signals and improving the accuracy of determining the time of their detection with SNR from 3 to 5 dB in the frequency band occupied by the detected radio signal.

To achieve the indicated technical result, in the method for detecting a plurality of narrow-band radio signals in a wide frequency band, analog-to-digital conversion of ballscrews, the accumulation of complex ballscrew samples and sequential sampling with ballscrews are performed, the amplitude spectrum of the generated sample with ballscrews is calculated, the detection threshold for each component is calculated amplitude spectrum as the sum of a smoothed spectrum estimate in a weighting window (ITS) sliding along the frequency axis and a function of their average standard deviation (RMS), comparing the amplitudes of the spectral components with the detection threshold, storing the results of comparing the amplitudes of the spectral components with the detection threshold, deciding on the detection of radio signals in the studied frequency band based on the majority processing of the stored results of comparing the amplitudes of the spectral components with the detection threshold, obtained in the analysis of six consecutively taken samples with ball screw samples, and the output of the detection results.

Common features of the prototype and the proposed method are analog-to-digital conversion of ballscrews, the accumulation of complex ballscrew samples and sequential sampling with ballscrews, the calculation of the amplitude spectrum of the sample with ballscrews, calculation of the detection threshold, the comparison of the amplitudes of the spectral components with the detection threshold, decision making about the detection of radio signals in the studied frequency band, the output of the detection results.

Distinctive features of the proposed method from the prototype are:

calculation of the detection threshold for each component of the amplitude spectrum as the sum of the smoothed spectrum estimate in the NWO and the function of their standard deviation;

saving the obtained results of comparing the amplitudes of the spectral components with the detection threshold;

making a decision on the detection of radio signals in the studied frequency band based on the majority processing of the stored results of comparing the amplitudes of the spectral components with the detection threshold value obtained by analyzing six consecutive samples with ball screw samples.

Thanks to a new set of essential features, the technical result - the reduction of the influence of STR, the transmission coefficient of the RPU path, the law of distribution of atmospheric, space interference and intrinsic noise of the RPU in the studied frequency band on the R _{PO of} narrow-band radio signals in the frequency band four times or more their width is manifested due to the fact that the detection threshold is calculated for each component of the amplitude spectrum as the sum of the smoothed spectrum estimate in the CBO and the function of their standard deviation. The obtained values of the detection threshold are determined only by the amplitude of the spectral components in the CBO and therefore do not depend on the STR outside the CBO. At the same time, a decrease in the width of the CBO leads to a decrease in the influence of the STR on the value of the detection threshold and, consequently, on the R _PO .

Thanks to a new set of essential features, the technical result is a decrease in the duration of detectable narrow-band radio signals and an increase in the accuracy of determining their detection time for SNR from 3 to 5 dB in the frequency band occupied by the detected radio signal, due to the fact that the decision to detect radio signals in the studied frequency band is made after majority processing of the stored results of comparing the amplitudes of the spectral components with the value of the detection threshold obtained at An analysis of six consecutively taken samples with ballscrew readings. Majority processing is implemented on the basis of the majority method [G. Nikitin Convolutional Codes: Tutorial. / SPbGUAP. St. Petersburg, 2001 .-- 80 pp., Ill. (p. 49)].

The analysis of the level of existing technology has made it possible to establish that there are no analogues that are characterized by a set of features identical to all the features of the claimed technical solution. This indicates the conformity of the claimed method to the condition of patentability "novelty." The search results for well-known solutions in this and related fields of technology in order to identify features that match the features of the claimed object that are different from the prototype showed that they do not follow explicitly from the prior art. The prior art also did not reveal the popularity of the impact provided by the essential features of the claimed invention, the transformations on the achievement of the specified technical result. Therefore, the claimed invention meets the condition of patentability "inventive step".

The method is illustrated by illustrations in which are presented:

figure 1 is a graph of the dependence of P _LT on the number of samples used in the majority processing of the results of comparing the amplitudes of the spectral components with the value of the detection threshold;

figure 2 - graphs of the dependencies R _{software of} radio signals on the number of spectral components in the CBO at different values of the coefficient used in the formula for calculating the detection threshold;

figure 3 is a graph of the dependence of P _LT from the coefficient used in the formula for calculating the detection threshold;

figure 4 - graphs of the dependencies R _{software of} radio signals on the number of spectral components in the CBO with a different number of samples used for majority processing;

5 is an amplitude spectrum of ballscrews and a detection threshold;

6 is a result of the detection of narrowband radio signals in a frequency band exceeding the width of their spectrum;

Fig.7 is a graph of the dependencies of R _software from loading the studied frequency band with narrow-band radio signals at SNR from 3 to 5 dB in the frequency band occupied by the detected radio signal.

A method for detecting a plurality of narrowband radio signals in a wide frequency band is as follows:

Stage 1. An analog-to-digital conversion of the ballscrew coming from the output of the broadband path of a panoramic radio receiver is performed.

The analog-to-digital conversion of ballscrew is implemented on the basis of the quadrature sampling method [Sergienko AB Digital Signal Processing: A Textbook for High Schools. 2nd ed. - SPb .: Peter, 2007 .-- 751 p .: ill. - ISBN 5-469-00816-9 (p. 168-170)]. As a result of quadrature sampling, complex ballscrew readings are formed. Using the method of quadrature sampling for analog-to-digital conversion of ballscrew allows you to lower the sampling frequency to a frequency corresponding to the upper limit of the passband of the radio receiving path.

To detect radio signals in the HF band with a spectrum width not exceeding 8 kHz (two standard telephone channels), as well as in the VHF band with a standard spectrum width of 20-30 kHz, this step can be implemented using the ARK-PR5 Argamak panoramic radio receiver with a 1 MHz radio channel passband in the HF band and 10 MHz in the VHF band [Rembovsky AM, Ashikhmin A.V., Kozmin V.A. Radio monitoring: tasks, methods, tools. / Edited by A.M. Rembovsky. - M .: Hot line - Telecom, 2006 .-- 492 p.: Ill. - ISBN 5-93517-326-3 (p. 88-92)] and a 14-bit high-speed data acquisition device ADC-14-200 with two synchronous channels, a sampling frequency of up to 200 MHz and a dynamic range of up to 87 dB [P. Dorofeev, Rudnev P. Modern high-speed ADCs with a large dynamic range. // Electronics: Science, Technology, Business. 2006. No. 4, p.23-25], capable of quadrature analog-to-digital conversion with a sampling frequency of 1 MHz and 10 MHz.

Stage 2. Accumulation of complex ball screw samples and sequential sampling with ball screw samples.

After the accumulation of the required number of samples, allowing to calculate the range of ballscrews, the sample with the samples is transmitted for further processing, implemented at stages 3-10. Simultaneously with the transfer of this sample for further processing, the accumulation of the next ball screw sample begins.

Stage 3. Calculation of the amplitude spectrum of the formed sample with ball screw samples.

The calculation of the values of the spectral components can be performed in various ways, for example, based on the discrete Fourier transform (DFT) [Richard Lyons. Digital Signal Processing: Second Edition. Per. from English - M .: Binom-Press LLC, 2007 - 656 p.: Ill. - ISBN 5-9518-0149-4 (p. 64)]:

where S [n] is the complex spectral readout of ballscrew;

n is the number of the complex spectral readout of ballscrew;

N is the number of samples in the sample and complex spectral samples in the ballscrew spectrum;

i - ball screw reference number;

X [i] - integrated ball screw reading;

exp - exponential function;

j is the imaginary unit.

For a given sampling frequency (f _D ), the number of samples in the ball screw sample determines its duration (τ _V ) and, therefore, the resolution of the DFT in frequency (Δf):

,

,

.

.

When determining the number of samples required for calculating the ballscrew spectrum, it is necessary to take into account that its duration should be M times shorter than the minimum duration of detected radio signals (τ _C ), where M is the number of samples of the stored results of the comparison of the amplitudes of the spectral components with the magnitude detection threshold.

Based on this, the maximum allowable number of samples in the ball screw sample (N _max ) is calculated by the formula:

To reduce the computational cost when calculating the range of ballscrews, the obtained value of N _{max is} reduced to the nearest number, determined by the formula:

where d is a positive integer greater than six.

For example, in the VHF band for detecting narrow-band radio signals with a standard spectrum width of 25 kHz and a duration of at least 0.66 ms with a SNR of 3 dB in the frequency band occupied by the detected radio signal, the bandwidth of the panoramic radio receiver 10 MHz and the same frequency value sampling of an analog-to-digital converter requires majority processing of six samples. Therefore, in accordance with (1), the maximum allowable size of one ball screw sample will be 1100 samples. The closest value that satisfies condition (2) is 1024 samples, which corresponds to a sample duration of 0.102 ms and a DFT resolution of 9.76 kHz. The duration of the six samples will be 0.612 ms, which does not exceed the minimum duration of the detected radio signals.

The detection of radio signals with the presented parameters is possible with a smaller number of samples in the sample, for example, with 512 samples. In this case, the sampling duration will decrease, which will increase the accuracy of determining the detection time of radio signals, however, the frequency resolution of the DFT will deteriorate by half (up to 19.53 kHz), which will reduce the accuracy of measuring such radio signal parameters as the spectrum width and carrier frequency.

An increase in the number of samples in a sample of ballscrews, for example, by a factor of two compared with a number satisfying condition (2), to 2048 will lead to an improvement in the frequency resolution (up to 4.88 kHz). However, the duration of one sample in this case will increase to 0.204 ms, which for a given signal duration will allow using only three samples of ballscrews for majority processing and will reduce the reliability of detection due to the high P _RT (Fig. 1).

To calculate the amplitude spectrum of ballscrews, the amplitudes of complex spectral readings are calculated [Richard Lyons. Digital Signal Processing: Second Edition. Per. from English - M .: Binom-Press LLC, 2007 - 656 p.: Ill. - ISBN 5-9518-0149-4. Page 66-67], the obtained values are stored in the array of amplitudes As:

where Re is the allocation function of the real part of the complex number;

Im is the function of extracting the imaginary part of a complex number.

Stage 4. Calculation of the detection threshold.

The detection threshold value is calculated for each component of the amplitude spectrum as the sum of the smoothed spectrum estimate in the CBO and the function of their standard deviation.

The width of the CBO (ΔF _CBO ) is calculated by the formula:

,

,

where k is the number of spectral components in the CBO.

The CBO shift step is equal to or a multiple of Δf. At each step, the threshold value for the central component of the amplitude spectrum in the CBO is calculated. As a result, in the calculation at each step, one point of the detection threshold is formed, the value of which is determined only by the amplitude k of the spectral components and therefore does not depend on the STR outside the CBO. The value of k is determined by the width of the spectrum of the detected radio signals, the resolution of the DFT and is selected at least twice as much as the number of spectral components in the frequency band of the detected radio signal. However, a significant increase in the width of the CBO leads to an increase in the dependence of the threshold value on the STR over the limits of the frequency band of the detected radio signal. The obtained detection threshold values are stored in the _software array G.

With this approach to calculating the detection threshold, its first element will correspond to the spectral component with number n _H , the value of which is determined by the formula:

,

,

and the last element is the spectral component with number n _P , the value of which is determined by the formula:

.

.

Therefore, the number of elements in the G array of _software will be less than the number of elements in the As array by N _NK elements. The value of N _{NK is} calculated by the formula:

.

.

To eliminate this drawback, before calculating the detection threshold, an AsN array is additionally formed, all elements of which are assigned the values of the first element of the As array, and also an AsK array, all elements of which are assigned the values of the last element of the As array. The number of elements in the arrays AsN and AsK (k _HK ) is calculated by the formula:

.

.

Then the general array A is formed according to the following rule:

where l is the number of the element in array A.

When using array A to calculate the detection threshold, its first point corresponds to the first component of the amplitude spectrum of the sample with ball screw samples, and the last point corresponds to the last component of the amplitude spectrum.

и функции от их СКО (σ_СВО) в соответствии с формулой:Each point of the G _software detection threshold is calculated as the sum of the smoothed estimate of the components of the amplitude spectrum in the CBO

and functions of their standard deviation (σ _CBO ) in accordance with the formula:

where r is the coefficient, the value of which depends on the required R _PO and the number of spectral components in the CBO.

The dependence of R _PO on k for various values of the coefficient r was established experimentally and presented in figure 2. These dependencies clearly demonstrate that an increase in r leads to a decrease in P _PO (figure 2) and P _LT (figure 3) for any width of the CBO.

For example, to detect radio signals whose spectrum is represented by no more than two spectral components, with an SNR of 3 dB in the frequency band occupied by the detected radio signal, in order to achieve the required R _software equal to 0.95, the value of the coefficient g is chosen equal to 0.3 . This choice is due to the following:

the minimum number of components of the amplitude spectrum in the CBO, at which a R _PO equal to 0.95 is achieved, is five, with smaller values of k of the required R _PO (Fig. 2) with SNR equal to 3 dB in the frequency band occupied by the detected radio signal, it is impossible to achieve ;

when choosing a larger value of the coefficient g, it is necessary to increase the number of components of the amplitude spectrum in the CBO (figure 2), which will increase the dependence of the detection threshold on the STR.

A smoothed estimate of the components of the amplitude spectrum in the NWO can be obtained by various methods, for example, as the mathematical expectation:

where t is the number of the spectral component in the CBO.

A value of t equal to zero corresponds to the number of the first component of the amplitude spectrum falling into the CBO at each step of the calculation, a value of t equal to k minus one represents the last component of the amplitude spectrum falling into the CBO at each step of the calculation.

также может рассчитываться как медианное значение спектральных составляющих в СВО, через выделение моды или другими методами.Value

can also be calculated as the median value of the spectral components in the CBO, through mode extraction or other methods.

The standard deviation of the components of the amplitude spectrum in the SVO is calculated as follows:

The range of ballscrews and the detection threshold calculated on the basis of (3) are presented in Fig. 5.

Step 5. Comparison of the amplitudes of the spectral components with the magnitude of the detection threshold.

A comparison of the amplitudes of the spectral components with the detection threshold is made for each sample of ballscrews. In this case, the amplitude of each spectral component is successively compared with the corresponding detection threshold. The comparison results are assigned to the elements of the array M _CP according to the following rule:

Using the results of comparison as the results of detection allows you to achieve a high value of R _software (figure 2). However, when this P _RT exceeds the value achieved when using the prototype in equal conditions.

It was experimentally established that when choosing to calculate the threshold value of the detection coefficient r equal to 0.3, P _LT reaches 0.8 (figure 3). The increase in the coefficient g to reduce P _{RT is} impractical, since it decreases P _ON (figure 2).

This determines the need for additional processing of the obtained comparison results, the purpose of which is to reduce P _RT , while maintaining the same P _PO value.

Step 6. Saving the results of comparing the amplitudes of the spectral components with the value of the detection threshold.

The elements of the array M _{CP are} recorded in a two-dimensional array with the results of the detection of M _OBN :

where m is the number of the processed sample, takes values from 1 to M.

Step 7. At this stage, the number of ball screw samples received from the output of the analog-to-digital converter for analysis is checked.

The number of ball screw samples used for majority processing is determined during the design of the detector and depends on the permissible value of P _LT . It was experimentally established that when choosing the coefficient r equal to 0.3, for calculating the detection threshold according to formula (3), the value of Р _ЛТ equal to 0.01 will be achieved by majority processing of the stored results of comparing the amplitudes of the spectral components with the value of the detection threshold obtained when analyzing six sequentially taken samples with ballscrew readings (Fig. 1). In this case, the decrease in P _software will be compensated by increasing the number of spectral components in the CBO to eleven (Fig. 4).

When using the prototype method, to achieve R _software equal to 0.95 for SNR equal to 3 dB in the frequency band occupied by the detected radio signal and P _LT equal to 0.01 is possible only by averaging the spectra over sixteen consecutive samples with ball screw samples .

Therefore, the use of the presented method allows to detect radio signals, the duration of which is 2.6 times less than that required for the prototype method in equal conditions of open source software. In addition, the use of majority processing reduces the error in determining the time of detection of radio signals. This ensures the reliability of detection is not worse than that of the prototype.

If the number of samples received from the output of the analog-to-digital converter for analysis is less than the value determined during the design of the detector, then steps 1-7 are performed. Otherwise, step 8 is performed.

Step 8. Majority processing of the stored results of comparing the amplitudes of the spectral components with the detection threshold

The majority processing of the stored results of comparing the amplitudes of the spectral components with the detection threshold value is carried out according to the following rule:

where R is an array with the result of majority processing of M samples.

Step 9. At this stage, a decision is made on the detection of radio signals in the studied frequency band and the output of the detection results.

The decision to detect a radio signal at a frequency corresponding to the number of an array element with the result of majority processing is taken if the value of this element is not equal to zero. Otherwise, a decision is made about the absence of a radio signal at a given frequency.

After the decision is made to detect radio signals at the appropriate frequencies, the results of the detection are output.

The detection results can be displayed, for example, on a graph whose abscissa axis corresponds to the frequency of the radio signal, the ordinate axis to its amplitude. To display the results of detection, the elements of the array of amplitudes As are multiplied by the corresponding elements of the array R with the results of the majority processing, the multiplication result is written to the output array W:

After that, the frequency value (f [n]) corresponding to the number of the array element is calculated:

where f _H is the initial frequency of the investigated frequency band.

Points with coordinates (f [n], W [n]) are sequentially displayed on the graph, a perpendicular is drawn from the point on the abscissa axis of the graph.

The result of the detection of narrow-band radio signals in a frequency band significantly exceeding the width of their spectrum by the proposed method is presented in Fig.6.

Step 10. An array with the results of comparing the amplitudes of the spectral components with the value of the detection threshold having the number m equal to zero is removed from the two-dimensional array M _OBN . The numbers of the remaining arrays are reduced by one. An array with the results of comparing the amplitudes of the spectral components with the detection threshold value obtained by processing the next sample is recorded in the freed memory region, and the transition to step 8 is performed.

Stages 2 ÷ 8, 10 can be implemented using a digital signal processing device based on a signal processor, for example ADSP-2181 [Valpa O.D. Development of devices based on digital signal processors from Analog Devices using Visual DSP ++. - M.: Hot line - Telecom, 2007 .-- 270 p. - ISBN 5-93517-342-5] or using an electronic computer (computer) with software that allows digital signal processing [Sergienko AB Digital Signal Processing: A Textbook for High Schools. 2nd ed. - SPb .: Peter, 2007 .-- 751 p .: ill. - ISBN 5-469-00816-9].

Stage 9 can be implemented on a computer with software that allows you to display graphical information.

The study of the feasibility of the proposed method was carried out on a personal computer through simulation.

To implement stage 1, an analog-to-digital conversion of ballscrews was performed in the frequency band 70-82 MHz.

To implement stages 2-10, software was developed in the C ++ programming language using the Visual Studio 2008 integrated development environment.

The ability to achieve the claimed technical result is confirmed experimentally. During the experiment, a radio signal with a spectral width of 25 kHz was detected when loading the investigated frequency band with narrow-band radio signals from 5 to 95% and SNR from 3 to 5 dB in the frequency band occupied by the detected radio signal. The value of the coefficient r is chosen equal to 0.3, the number of spectral components in the CBO equal to 11.

In the course of the experiments, the dependences of P _software on the load of the investigated frequency band by narrow-band radio signals at SNR from 3 to 5 dB in the frequency band occupied by the detected radio signal were obtained (Fig. 7). These dependencies confirm the achievement of the claimed technical result - a decrease in the influence of STR on R _{software of} narrow-band radio signals in the frequency band exceeding the width of their spectrum by four times or more.

The value of the detection threshold is calculated for each component of the amplitude spectrum and depends only on the eleven spectral components falling into the CBO, which confirms the achievement of the claimed technical result - reducing the influence of the transmission coefficient of the RPU path and the distribution law of atmospheric, space interference and intrinsic noise of the RPU in the studied frequency band by R _PO narrowband radio signals in a frequency band exceeding the width of their spectrum by four times or more.

When conducting this experiment, it was found that P _RT does not exceed 0.01 during the majority processing of six samples with the results of comparing the amplitudes of the spectral components with the value of the detection threshold, which confirms the achievement of the claimed technical result - reducing the duration of detectable narrow-band radio signals and improving the accuracy of determining the time of their detection with SNR from 3 to 5 dB in the frequency band occupied by the detected radio signal.

Claims (1)

A method for detecting a plurality of narrow-band radio signals in a wide frequency band, including analog-to-digital conversion of a wide-band input process (ballscrew), accumulating complex ballscrew samples and sequential sampling with ballscrews, calculating the amplitude spectrum of a sampled with ballscrews, calculating the detection threshold, comparing amplitudes spectral components with a detection threshold, a decision on the detection of radio signals in the studied frequency band, output result in detection, characterized in that the detection threshold value is calculated for each component of the amplitude spectrum as the sum of the smoothed spectrum estimate in the sliding window along the frequency axis and the function of their standard deviation, the results of comparing the amplitudes of the spectral components with the detection threshold value are saved, and the decision about the detection of radio signals in the studied frequency band is carried out on the basis of majority processing of the stored results of comparing the spectral amplitudes with stavlyayuschih with a detection threshold value obtained in the analysis of the six samples successively taken samples with leadscrew.

Images