Information-theoretic algorithm for waveform optimization within ultra wideband cognitive radar network

A novel information-theoretic approach for designing the excitation ultra wideband (UWB) waveforms within a cognitive radar network is developed. This method utilizes the mutual information (MI) between subsequent radar returns to extract desired information from the radar scene. With this approach, the radar system constantly learns about its surroundings and adopts its operational mode accordingly based upon the MI minimization criterion. Subsequently, the positioning algorithm makes use of this information about the radar scene to generate more accurate location estimates. Numerical results demonstrate an improvement in the probability of target detection even at low values of receive signal-to-noise ratio (SNR). The proposed algorithm also promises a better delay-Doppler resolution of the target, which can be analyzed through the radar ambiguity function (AF). Simulation data show an improvement in the target discrimination ability in the presence of noise and clutter.