The development will reduce the probability of an error when transmitting information using telecommunications equipment.

Today, one of the most important and rapidly developing scientific and technological fields is the field of telecommunications - systems that can provide uninterrupted and high–quality communication between individuals and countries anywhere on the planet, underwater and in space. In addition, they provide quick access to telephony, the Internet, television, online banking and many other consumer services.

At the same time, the number of users of communication systems increases every year, and, accordingly, the amount of data transmitted by them increases. On the one hand, this process leads to the scaling of the telecommunications infrastructure. And on the other hand, the number of signals simultaneously transmitted in existing information transmission channels increases – this increases the likelihood of distortion and errors in the data exchange process.

“We have developed a neural network model that allows us to automatically select a set of radio signals with the most effective data transmission parameters for a specific telecommunications system, based on the characteristics of the communication channel that is used in it,” explains Alexander Sergienko, Deputy Head of the Department of Theoretical Foundations of Radio Engineering at St. Petersburg State Technical University “LETI” for academic work.

The created neural network is a virtual communication system in which the processes occurring with signals during the transmission of information are simulated. The model solves the optimization problem, that is, it selects such a set of signals that, after their distortion by the communication channel, can be most reliably distinguished from each other.

In the process of optimization, the neural network takes into account a number of parameters for various telecommunication systems, the most important of which are the data transfer rate (spectral efficiency) and the signal level required to work with a given quality (energy efficiency). In communication systems, these parameters are conflicting: increasing spectral efficiency allows the communication system to transmit more data, but reduces the likelihood that information will reach the user without errors, and vice versa.

“Thus, with the help of our optimization model, it is possible to create sets of signals for existing communication systems, increasing the efficiency of their work in terms of the ratio of speed and reliability of information transmission,” says Alexander Sergienko.

The research was carried out within the framework of the work of SPbSETU "LETI" on the implementation of the "Priority-2030" program.