Radar Emitter Signal Detection with Time-frequency Image Processing and Convolutional Neural Network

In this paper, we propose a deep convolutional
neural network (CNN) based automatic detection algorithm for
recognizing radar emitter signals. The algorithm leverages on the
structure estimation power of deep CNN and the capability of
time-frequency image processing for radio signal representation.
We transform raw radio signals into time-frequency image using
the Choi-Williams distribution function. We compare the proposed
method with Belief Propagation (BP) and Support Vector
Machine (SVM) based methods in terms of recognition accuracy
versus signal-to-noise-ratio. The experiments demonstrate that
the proposed CNN network with time-frequency image processing
achieves very competitive results on the testing datasets.