O. S. Litvinov – Dr.Sc. (Phys.-Math.), Professor,

Department of Physics, Bauman Moscow State Technical University

E-mail: oleglitv@mtu-net.ru

D. V. Murodyants – Student,

Department of Physics, Bauman Moscow State Technical University

E-mail: dmitriy1994.94@mail.ru

V. S. Boruta – Ph.D. (Phys.-Math.), Associate Professor,

Department of Physics, Bauman Moscow State Technical University

B. E. Vintaykin – Dr.Sc. (Phys.-Math.), Professor,

Department of Physics, Bauman Moscow State Technical University

The aim of this work is to create an adaptive antenna array (AAA) with neural network phase control, which maximizes the ratio of the power of the useful signal to noise at the output with minimal loss of the useful signal power. The choice of a neural network as a control unit was determined by difficulty of solving the problem of phase control of the AAA, which does not have a known analytical solution. The tasks of the work also included conducting studies of various noise suppression characteristics in an adaptive antenna array on a neural network phase control, depending on the spectrum of the incident noise. In this paper, we considered three types of the incident noise spectrum – triangular, rectangular, and Gaussian. The main characteristics were considered: the dependence of the ratio of the power of the useful signal to noise at the output on the angle of arrival of broadband interference, as well as the dependence of the ratio of the power of the useful signal at the input to the power of the useful signal at the AAA output of the neural network phase control on the angle of arrival of the noise. The successful operation of AAA on neural network phase control in suppressing interference of various spectral composition, which was trained at the same time to suppress one monochromatic interference, was demonstrated. Based on the obtained dependences for signal-to-noise ratio and the ratio of the powers of the useful signal at the input to the AAA output, it is difficult to process signals of high spatial frequency, due to which the angle of arrival of the interference should not be close to the sliding angle, as well as to the normal angle, since will match the useful signal.

1. Litvinov O.S., Murod'yants D.V., Boruta V.S., Vintajkin B.E. Podavlenie shumovykh pomekh v adaptivnykh antennykh reshetkakh na osnove nejrosetevykh algoritmov. Antenny. 2018. № 2. S. 40–44. [in Russian]
2. Pistol'kors A.A., Litvinov O.S. Vvedenie v teoriyu adaptivnykh antenn. M.: Nauka. 1991.
3. Khajkin S. Nejronnye seti: Polnyj kurs. Izd. 2-e. M.: ID «Vil'yams». 2008.
4. Uossermen F. Nejrokomp'yuternaya tekhnika: Teoriya i praktika. M.: Mir. 1992.
5. Acharya N., Acharya S., Panda S., Nanda P. An artificial neural network model for a diesel engine fuelled with mahua biodiesel. Computational Intelligence in Data Mining: Proceedings of the International Conference on CIDM. 2017. P. 193–201. DOI: 10.1007/978981-10-3874-7_19
6. Liu G., Jia L., Kong B., Guan K., Zhang H. Artificial neural network application to study quantitative relationship between silicide and fracture toughness of Nb-Si alloys. Materials & Design. 2017. V. 129. P. 210–218.
7. Cybenko G.V. Approximation by superpositions of a sigmoidal function. Mathematics of control, signals and systems. 1989. V. 2. P. 303–314.
8. Siegelmann H.T., Sontag E.D. Analog computation via neural networks. Theoretical Computer Science. 1994. V. 131. № 2. P. 331–360.
9. Islam Md.S., Mohandes M., Rehman Sh. Vertical extrapolation of wind speed using artificial neural network hybrid system. Neural Computing and Applications. 2017. V. 28. P. 2351–2361.
10. Zinov'ev A.Yu. Vizualizatsiya mnogomernykh dannykh. Krasnoyarsk: Izd-vo Krasnoyarskogo gos. tekhn. un-ta. 2000.
11. Zelkin E.G., Kravchenko V.F. Sintez antenn na osnove atomarnykh funktsij. M.: Radiotekhnika. 2003.
12. Graves A., Schmidhuber J. Offline handwriting recognition with multidimensional recurrent neural networks. 2009. V. 31. P. 545–552.