Enhanced LPI Waveform Representation by Ambiguity-Domain Elliptical Gaussian Filtering

Abstract

ESM systems need to use sophisticated signal processing, such as time-frequency representation techniques, for the interception of nonstationary LPI radar signals. In this study, an adaptive filtering technique using an ambiguity-domain elliptical Gaussian kernel is proposed to increase the readability of pseudo-Wigner-Ville distribution based representations for LPI waveform parameter extraction purposes. The complexity and information content of the outputs obtained by the proposed method are evaluated by objective criteria, such as ratio of norms, Rényi entropy, and Jubisa measure. The results quantify efficient filtering performance under severe SNR conditions with lower computational complexity.