Quasi-orthogonal wideband radar waveforms based on chaotic systems

Abstract

With the development of A/D converters possessing sufficiently high sampling rates, it is now feasible to use arbitrary, wideband waveforms in radar applications. Large sets of quasi-orthogonal, wideband waveforms can be generated so that multiple radars can simultaneously operate in the same frequency band. Each individual radar receiver can process its own return as well as the orthogonal returns from the other radars, which opens the possibility for developing algorithms that combine data from multiple radar channels. Due to the random nature of chaotic signals, it is possible to develop a procedure for generating large sets (> 50) of quasi-orthogonal radar waveforms using deterministic chaos. Deterministic chaos is defined as a bounded, aperiodic flow with a sensitive dependence on initial conditions. There are many different types of chaotic systems. In this thesis, waveforms will be generated from the well-studied Lorenz system. Each waveform from the Lorenz system can be fully characterized by three parameters (o, b, and r) and a set of initial conditions, (xo, yo, zo). The particular parameter values greatly affect quality of the Lorenz waveform as quasi-orthogonal radar waveform. Therefore, this thesis conducts a parameter study to quantify how the parameters affect various radar waveform metrics. Additionally, this thesis proposes a procedure for modifying the Lorenz waveform in order to improve its performance on these metrics.