DEVELOPMENT OF A HARDWARE-SOFTWARE SYSTEM FOR IDENTIFICATION OF DEVICES IN WIRED COMMUNICATION LINES

The article addresses hidden attachments in wired lines that remain invisible to logical monitoring and can cause information leakage and service disruption. The aim is to develop a device identification system that detects interventions and assigns the connected object to a class using electromagnetic signal features while keeping the channel in service. The approach relies on bench trials on typical lines, a combination of active probing and passive observation, acquisition of a reference state before connection and an observed state after connection, construction of a time-frequency signal portrait, adaptive amplitude thresholding, normalization for noise and temperature, a fixed line profile, and matching against a reference library. The work identifies informative spectral bands and temporal windows, demonstrates separability of representative intervention scenarios, and confirms feature reproducibility under changing external conditions, which follows from combining probing with passive capture and applying a unified processing rule. The conclusion is that physical-layer control can be integrated into routine operations with operator-readable decisions and transferable settings, making the system suitable for monitoring in corporate and industrial networks and for technical information protection.