RADIOELECTRONICS
DFTN matrix optimization to increase the algorithm noise immunity
Mikhail G. Bakulin, Vitaly B. Kreyndelin
Moscow Technical University of Communications and Informatics, https://mtuci.ru/
Moscow 111024, Russian Federation
Email: m.g.bakulin@gmail.com, vitkrend@gmail.com
Andrey A. Reznev
Scientific-Technical Center "Atlas", https://www.stcnet.ru/
Moscow 127018, Russian Federation
Email: andrey.reznev@gmail.com
Received April 07, 2025, peer-reviewed April 14, 2025, accepted April 17, 2025, published August 14, 2025
Abstract: The article is devoted to methods for optimizing the transformation matrix for an algorithm for discrete signal transmission with a speed above the Nyquist limit, which makes it possible to increase the spectral efficiency of communication systems. The transmission of information at speeds exceeding the Nyquist limit makes it possible to increase the spectral efficiency of communication systems by reducing the protective transmission intervals with a controlled amount of intersymbol interference. For modern communication systems, it is necessary to use a discrete version of this algorithm. Discretization is achieved by organizing the transmission of modulated symbols over a certain number of time intervals. The ratio between a large number of symbols and a smaller number of intervals affects both the spectral efficiency of the communication system and its noise immunity. The presented relationship is described by a transformation matrix, the type and dimension of which affect the characteristics of the communication system. The authors propose an approach related to the formation of the transformation matrix, its individual elements – the rotation of each element of the matrix by a certain angle and the forced addition of an excess symbol. An algorithm for optimizing the transformation matrix based on the gradient method is also presented. The obtained analytical conclusions are confirmed by the results of simulation modeling. Thus, the energy gains of communication systems with optimal matrices range from 0.14 to 2.15 dB, with different ratios between the number of modulated symbols and transmission samples. The presented results demonstrate that optimization of the transformation matrix, which can be performed in advance, increases the noise immunity of communication systems compared to the original matrices. In addition, the proposed methods for forming matrix elements also increase the noise immunity of the communication system.
Keywords: spectral efficiency, FTN, DFTN, noise immunity, constellation rotation, transformation matrix
UDC 621.396
RENSIT, 2025, 17(4):419-430e
DOI: 10.17725/j.rensit.2025.17.419
Full-text electronic version of this article - web site http://en.rensit.ru/vypuski/article/689/17(4)419-430e.pdf