Adaptive waveform design for target classification

For active sensors, waveform/signal optimization is of great importance to improve system performance. In this paper, the adaptive waveform parameter is designed to improve the classification performance by minimizing the Bayesian error probability for the optimal decision of a symmetric binary hypothesis testing problem. It is well known that the probability of error can be bounded by the Chernoff divergence between the distributions of the two hypotheses. Therefore, by maximizing the Chernoff divergence between the two distributions of the hypotheses, the optimal waveform parameter is obtained to enhance the classification performance. Simulation results prove that the adaptive optimal waveform outperforms the fixed parameter waveform.