Chaotic and chaotic-based frequency-modulated signals for radar imaging.

摘要

This thesis presents an analysis of a chaotic and chaotic-based frequency-modulated signals, and demonstrates the possibility of their use in wideband radar imaging systems. The signals resolution capability and sidelobe distribution on the range-Doppler plane are assessed by the inspection of their ambiguity surface. Four deterministic, bounded, non-linear, iterated maps are considered. Statistically independent samples with invariant probability density function are obtained by randomizing the initial condition of each map. The resulting sequences, which have broadband frequency representations, are also used to construct wideband, stochastic frequency-modulated signals. These chaotic and chaotic-based frequency-modulated signals are ergodic and stationary. The autocorrelation, spectrum, and the ambiguity surface associated with the signals are characterized, as well as their chaotic behaviour using the Lyapunov exponent and correlation dimension. The ambiguity surface of an FM signal generated via a chaotic map with uniform sample distribution is demonstrated to be comparable to the ambiguity function of a random FM signal and superior to that of a linear frequency-modulated chirp. Finally the proposed new radar signals are compared to other traditional waveforms used in radar systems in terms of resolution, electronic counter-countermeasures and multipath performance.