Based Chaotic Radar Waveform Design Studies

Due to the sensitivity to initial condition, controllability, aperiodicity, noise-like character, favorable autocorrelation performance, thumbtack-shaped ambiguity function, continuous wideband spectrum and low probability of intercept (LPI), chaotic signals draw much attention in radar domain. The widely studied chaotic frequency modulated (FM) signals and phase modulated(PM) signals have been applied and perform well in the resolution of distance, radar imaging and capability of ECCM. However, chaotic FM signals are mostly based on discrete chaotic systems , the autocorrelation sidelobe of signals needs to be further suppressed. Bernulli chaotic FM signal is one of the typical chaotic waveform with sidelobe about -28dB, it could lead to loss of signal-noise-ratio (SNR) while applying some kind of sidelobe suppression algorithms; Logistic binary-phased code is the most common binary-phased code, but it needs to reduce the peak of cross-correlation function and the fluctuation of the in-band power spectrum; meanwhile, not much has been done to study the LPI property of chaotic signals. In order to resolve these problems, the following innovations have been discussed in the dissertation,Firstly, the characteristics of discrete chaotic FM signals are analyzed, including ergodicity, auto-correlation, cross-correlation, spectrum and ambiguity function. Based on this, a new chaotic FM signal which is modulated by the variable of continuous chaotic system after optimizing initial condition value is proposed, as a result, Lorenz FM signal is produced which could remarkably reduce the sidelobe level to -39.53dB while maintaining the SNR. After comparing with discrete chaotic FM signals, the conclusion can be drawn that the Lorenz FM signal has better sidelobe suppression performance.Secondly, several kinds of binary-phased codes are simulated and compared, among which Chebyshev binary-phased code is the best, it outperforms Logistic binary-phased code in the aspects of autocorrelation, cross-correlation and spectrum property. Then, how initial value, iteration times, starting point and length of the binary-phased code affect the auto-correlation sidelobe is investigated and summarized. At last, chaotic radar signals are compared with LFM in different noisy environments, simulation results demonstrate that the chaotic signal always has better LPI performance than LFM signal.