CNN based Modulation classifier and Radio Fingerprinting for Electronic Warfare Systems

The discernment of individual electronic radiators is facilitated by their unique emission patterns, enabling the potential identification of transmitters such as radios and radars based on waveform characteristics and modulation, as depicted in constellation diagrams. The advancement of Software-Defined Radio (SDR) technology has notably augmented the capacity to intercept transmissions. By harnessing artificial intelligence, one can scrutinize the In-phase and Quadrature (IQ) data alongside waterfall diagrams associated with each transmission to validate transmitter identification. Consequently, this can unveil the Electronic Order of Battle (EOB) of the adversary within a given scenario. This paper delineates two methodologies for transmitter identification and introduces the notion of a hybrid model to further refine the accuracy and expeditiousness of this process.