Method for determining intentional interference type in conditions of uncertainty of interference environment based on known models

The study aimed to theoretically substantiate a method for the identification of intentional interference types in an uncertain interference environment for software-defined radio systems by using known interference models and formalised analysis of the communication channel state vector. The methodological basis included window calculation of the matched feature vector, their normalisation, signature coding, metric and probabilistic comparison with a template library, threshold rejection of uncertain cases, and model-simulation verification in the streaming architecture of software-defined radio. As a result, a formalised feature space and signature profiles of five basic classes of interference were formed: impulse, broadband noise, harmonic, modulation, and lowvisibility – as stable multi-parameter patterns of common signal indicator changes that ensure class separation in different signal-to-noise ratio modes, bit error probability variability, and spectral power density morphology, and are reproduced in repeated runs with unchanged parameters. The method is implemented in the form of a computational pipeline with a sequence of stages “state vector → event detection → segmentation → signature description → dimensionality reduction → comparison → decision”, where resource-intensive operations are activated only in the presence of an event trigger, and the interfaces between the blocks record a transition from observations to a compact impact signature. The resulting classification rules with membership thresholds and a controlled rejection procedure have been formulated, which transfers new signatures to the accumulation mode and incremental supplementation of the database with subsequent confirmation of profile stability on repeated observations. Simulation testing on a series of parameterised scenarios showed the reproducibility of signature profiles and classification decisions under fixed settings. The practical significance of the study results is determined by the possibility of implementing the method by developers and integrators of embedded software-controlled radio systems and radio monitoring systems for streaming determination of the type of intentional interference based on signature profiles with threshold rejection of unknown cases and controlled replenishment of the signature library