Application of Pseudorandom Number Generators in Machine Learning

This paper presents a comprehensive study of methods for detecting and localizing electronic warfare (EW) sources under complex conditions, which include the influence of intensive intentional jamming, multipath signal propagation, time-varying emission parameters, and dynamic movement of the sources. The relevance of the research is driven by the growing role of EW means in modern armed conflicts and the need to ensure timely detection and neutralization of such threats.

The classification of EW sources based on spectral, energy, and spatial characteristics is considered, along with an analysis of their impact on communication, radar, and navigation systems. Traditional methods for determining the coordinates of sources-such as amplitude, phase, TDOA (Time Difference of Arrival), and FDOA (Frequency Difference of Arrival)-are examined in detail, as well as modern approaches that rely on adaptive antenna arrays, spectral analysis, and artificial intelligence algorithms for signal identification.

Special attention is given to algorithms for improving localization accuracy in noisy and dynamic environments, including adaptive filtering methods, multiposition signal processing (data fusion), optimization of sensor network configurations, and trajectory prediction of moving objects. Results of computer simulations are presented, with error assessments for various application scenarios and different interference levels.

The proposed approaches can be applied in advanced electronic intelligence systems and automated EW countermeasure complexes to enhance the efficiency of real-time detection, identification, and localization of emission sources.