Research On Detection And Parameter Estimation Of LFM Signals In Low SNR

On basis of the project “Signal reconnaissance and processing technology of ×× radiation source exploiting RSA(Real-time Spectrum Analysis)”, research on reconnaissance processing technology of radar signals is undertaken. Characteristics of linear frequency modulation(LFM) signals are studied, detection methods of which are referred to as well as approaches of parameter estimation. The entire paper is mainly composed of three parts:To start with, time-frequency features are analyzed with time-domain methods while frequency features are analyzed with frequency-domain methods. Moreover, time-frequency features are given by making use of Short-Time Fourier Transform(STFT), Wavelet transform, Wigner-Ville Distribution and pertinent derived ways and means.Second, Constant False Alarm Rate(CFAR) detection of LFM signals is studied. Basic rationale and development status quo of CFAR processing are introduced, module analysis of several logical threshold logics afterwards. And then a LFM signal detection algorithm based on STFT and frequency-domain CFAR is proposed. Input signals are short-time Fourier transformed to achieve coherent integration of frequency-shift sample sequences with complex envelopes, modular square of which are to be GOSBOS-CFAR detected in frequency domain. Simulation results show that the algorithm has fine detection performance and the GOSBOS logic outweighs some classical thresholds in controlling false alarm, thus making it suitable to detect LFM signals in low SNR(Signal-to-Noise Ratio).The last part focuses on parameter estimation of LFM signals. Parameters to be estimated involve pulse time of arrival(TOA), pulse recurrence interval(PRI), pulse width, bandwidth, starting frequency, cut-off frequency, frequency modulation slope. As an integral algorithm which is an access to integration of radar signals, Radon transform is specifically introduced in this part. Furthermore, a detection and parameter estimation method based on STFT and Radon transform is proposed. Right after the process that time-frequency information of the signal is achieved by STFT, lines are to be detected with subsection Radon transform extracting parameters such as bandwidth etc. Detection performance and estimation error are analyzed in conclusion.