Radar Emitter Identification using Signal Noise and Power Spectrum Analysis in Deep Learning
Akhila Madhu, Prajeesha Prajeesha, Abhijit S Kulkarni · 2022
Specific emitter identification (SEI) is a vital function of the electronic radar warfare support system. The challenge emphasizes recognizing and locating unique transmitters, avoiding potential threats, and preparing a countermeasure. Unlike the analog transient parameters, radar fingerprinting is effectively possible with digitized feature extraction and image processing methods. Our novel approach utilizes the power spectrum and signals noise to efficiently work on a large image dataset using deep learning techniques. The convolution neural network is tested on various time-frequency estimators to yield the most accurate results. Based on signal-to-noise ratios, radar emitters get distinguished using the ablest estimator.