Modulation Pattern Recognition And Parameter Estimation For Ground Wave Radiation Sources

LFM signal and PCM signal are the two most frequently used format of ground wave signals. As a typical nonstationary and low interception probability signal, LFM signal has a wide usage in different domains, especially in the domain of electronic warfare. It's meaningful to detect and estimate the parameter of LFM signal in the low SNR background. This paper is mainly based on the theory of time-frequency transform to estimate the parameter of LFM signal, and dechirping is also used to deal with the problem. At the end of the paper, we discuss some parameter estimation methods to PCM signals.After the introduction of some formats and summarizations of time-frequency transform to LFM signal is exhibited, the paper discusses the characteristics of the formats in the process of parameter estimation to single and multi-LFM signals. Then the generalized Cohen function is given, and the intermodulation reduction results of fixed kernel to single and multi-LFM signals are also delivered. Then I combine the line character of LFM in the time-frequency domain and the estimation merits of Radon transform to estimate the parameters of LFM signal. The models are Radon-WVD, Radon-STFT, and Radon-Ambiguity accordingly. Based on the inspiration of Radon-Ambiguity transform, the paper puts forward an adaptation kernel design method ground on the LFM parameters, and make a comparation between the results of the fixed kernel and the adaptation kernel. At the end of the paper, I take a view to the time domain dechirping method to LFM signals and estimation to PCM signals based on the sliding-window technic. The time domain methods have a better time-saving character compared to the time-frequency transform, and take better practical values.Finally, I make a conclusion to the whole paper, and bring forward some flaws that should be improved in future work during the process of parameter estimation to LFM and PCM signals.