PRI Modulation Classification in EW Systems Using Deep Learning

PRI Modulation Classification in EW Systems Using Deep Learning

연구 분야: Infrastructure

학회: International Conference on Emerging Global Trends in Engineering and Technology

The analysis of radar signals is a critical task in Electronic Warfare (EW) environments and decides the nature of counter employments. For an Electronic Support (ES) system, the challenge is to efficiently recognize the source of the threat radiation. In dense EW situations, detection of Pulse Repetition Interval (PRI) modulation modes of radar signal significantly aids the manifestation of emitter in the process of radar emitter recognition. Developments in Artificial Intelligence (AI) methods suggest that this emerging technology can be very effective for such purposes. In this direction, an automatic approach for recognizing eight kinds of complex PRI modulation types based on Continuous Wavelet Transform (CWT) and Convolutional Neural Network (CNN) is proposed. The CWT is used to decompose the PRI modulation sequence to obtain different time–frequency components, and the CNN is used to extract features from the 2D scalogram composed of temporal and spectral elements for deriving appropriate class decisions. The simulation result shows that the proposed method not only enhances the performance but is also robust in an environment with noisy content. The recognition accuracy is 98.2% with 30% spurious pulses in the simulation environment.