Research On Methods For Time-Frequency Analysis Of Radar Emitter Signals

Radar emitters signal recognition is an important component of signal processingin ELINT and ESM systems. It is also the precondition and foundation of electronicinterfering. The state of the art of radar emitter signal recognition corresponds to thetechnical merit of electronic reconnaissance equipment. Since the time-frequencyanalysis characterizes signal energy or strength in both time and frequency domains,namely, time-frequency analysis describes the spectral of the signal changes over time,it’s widely applied in radar emitter signal recognition. In this paper, we use differenttime-frequency analysis methods to deal with the denoising and multipath detectionproblems in radar emitter signal recognition. The research finding is very helpful toprovide valid signal information for the further feature extraction and classification.First of all, some basic concepts of time-frequency analysis are introduced. Theshort time Fourier transform (STFT), fraction transform (FRFT) and other commontime-frequency analysis methods are discussed in term of their advantages anddisadvantages.Secondly, for the denoising of radar emitter signal, a new method based on theimproved threshold of wavelet transform (WT) is proposed. The basic concept ofwavelet transform is introduced firstly, followed by the discussion of the traditionalthreshold denoising approach and great value method are introduced. By improving thethreshold method in WT, a better denoising result is obtained. The experiment resultsshow that the signal to noise ratio can be improved greatly.Finally, for the multipath detection in radar emitter recognition, local polynomialFourier transform (LPFT) is introduced. By deriving the time domain mathematicalmodel of multipath signals, the novel multipath detection algorithm based on LPFT isproposed. The simulation results verificate the effectiveness of the algorithm.