Permutation entropy for signal analysis: A case study of synthetic aperture radar imagery

Shannon entropy is a powerful tool for signal analysis. However, because it is based on a reductionist worldview, on its own, Shannon entropy cannot properly handle `temporal dynamics' pertaining to nonlinear and nonstationary processes. To remedy this, an extension of Shannon entropy, known as permutation entropy (PE), has been proposed in the literature. In this paper, the utility of PE for signal analysis is demonstrated on a comprehensive and real-world synthetic aperture radar dataset. When compared to conventional methods for signal analysis, the results convey the statistical significance of PE in capturing the dynamics between the constituent values of the signal.