Advanced signal processing solutions for ATR and spectrum sharing in distributed radar systems

  • Domenico Gaglione

Student thesis: Doctoral Thesis

Abstract

This Thesis presents advanced signal processing solutions for Automatic Target Recognition (ATR) operations and for spectrum sharing in distributed radar systems. Two Synthetic Aperture Radar (SAR) ATR algorithms are described for full- and single-polarimetric images, and tested on the GOTCHA and the MSTAR datasets. The first one exploits the Krogager polarimetric decomposition in order to enhance peculiar scattering mechanisms from manmade targets, used in combination with the pseudo-Zernike image moments. The second algorithm employs the Krawtchouk image moments, that, being discrete defined, provide better representations of targets' details. The proposed image moments based framework can be extended to the availability of several images from multiple sensors through the implementation of a simple fusion rule. A model-based micro-Doppler algorithm is developed for the identification of helicopters. The approach relies on the proposed sparse representation of the signal scattered from the helicopter's rotor and received by the radar. Such a sparse representation is obtained through the application of a greedy sparse recovery framework, with the goal of estimating the number, the length and the rotation speed of the blades, parameters that are peculiar for each helicopter's model. The algorithm is extended to deal with the identification of multiple helicopters flying in formation that cannot be resolved in another domain. Moreover, a fusion rule is presented to integrate the results of the identification performed from several sensors in a distributed radar system. Tests performed both on simulated signals and on real signals acquired from a scale model of a helicopter, confirm the validity of the algorithm. Finally, a waveform design framework for joint radar-communication systems is presented. The waveform is composed by quasi-orthogonal chirp sub-carriers generated through the Fractional Fourier Transform (FrFT), with the aim of preserving the radar performance of a typical Linear Frequency Modulated (LFM) pulse while embedding data to be sent to a cooperative system. Techniques aimed at optimise the design parameters and mitigate the Inter-Carrier Interference (ICI) caused by the quasiorthogonality of the chirp sub-carriers are also described. The FrFT based waveform is extensively tested and compared with Orthogonal Frequency Division Multiplexing (OFDM) and LFM waveforms, in order to assess both its radar and communication performance.
Date of Award26 May 2017
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde
SupervisorJohn Soraghan (Supervisor) & Stephan Weiss (Supervisor)

Cite this

'