Fourier independent component analysis of radar micro-Doppler features

Addabbo, P. and Clemente, C. and Ullo, S. L.; (2017) Fourier independent component analysis of radar micro-Doppler features. In: 4th IEEE International Workshop on Metrology for AeroSpace, MetroAeroSpace 2017 - Proceedings. IEEE, ITA, pp. 45-49. ISBN 9781509042340 (https://doi.org/10.1109/MetroAeroSpace.2017.799952...)

[thumbnail of Addabbo-etal-MetroAeroSpace-2017-Fourier-independent-component-analysis-of-radar-micro-Doppler-features]
Preview
Text. Filename: Addabbo_etal_MetroAeroSpace_2017_Fourier_independent_component_analysis_of_radar_micro_Doppler_features.pdf
Accepted Author Manuscript

Download (1MB)| Preview

Abstract

The capability of discriminating radar targets exhibiting multiple moving parts has become of great interest for both aerospace and ground-based target recognition and analysis. In particular, helicopters and other targets with rotors, as for instance miniature Unmanned Aerial Vehicles, exhibit peculiar characteristics in the radar return that can be used for their recognition. In this paper a novel algorithm to address the problem of micro-Doppler signature unmixing is proposed, exploiting the signal separation capabilities of the Independent Component Analysis (ICA). The core of the algorithm is represented precisely by the use of the ICA procedure, that has been already proved to be a very effective technique for separating hidden information in mixtures of observations. ICA has been successfully employed in several applications such as wireless communications, radar beamforming, trace-gases unmixing and medical imaging processing. The helicopter's rotor blade signature unmixing from a multi-static radar system is considered as case study and results obtained through the application of ICA to simulated multi-component micro-Doppler signatures show the capability of the proposed approach to successfully accomplish the unmixing operation.