Use this URL to cite or link to this record in EThOS:
Title: Distributed sensing for MIMO radar systems
Author: Ilioudis, Christos V.
ISNI:       0000 0004 7960 8659
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
The research presented in this thesis deals with the concepts of distributed sensing for multiple-input multiple-output (MIMO) radar systems and important signal processing algorithms with regard to multiple sensing optimisations. These novel algorithms include an edge detection scheme based on the phase stretch transform (PST) for synthetic aperture radar (SAR) imaging systems, the application of the fractional Fourier transform (FrFT) in generating new waveform libraries and the synthesis of a generalised MIMO ambiguity function (AF) based on the Kullback-Leibler divergence (KLD). In particular, a new edge detection algorithm for SAR images is proposed. This method is an enhanced scheme that is based on the phase stretch transform (PST). The high-accuracy of the presented edge detection method is tested and verified experimentally using two SAR image datasets. Experimental results show that thresholding and further morphological operation leads in excellent edge extraction despite the noise embedded into the image. Including PST into the structure of the edge detection algorithm is proved to be very advantageous, since the efficiency in edge determining could be improved by means of tuning the strength and wrap parameters of PST phase kernel. It is shown that the proposed method is very effective and capable to remove embedded noise and introduced artefacts even from image parts corresponding to the surface of the sea. A novel waveform design scheme is proposed to create waveform libraries employing the FrFT. Additionally an efficient algorithm based on a modified Gerchberg-Saxton algorithm (MGSA) is developed to reconstruct the proposed fractional waveform libraries under constant envelope (CE) constrain. This efficient technique is capable of generating novel libraries of phase-coded waveforms through FrFT and optimise the signal retrieval, while the signal waveforms retain their constant modulus. Specifically, the reconstruction of sequences from the FrFTbased waveforms is achieved by means of the error reduction algorithm (ERA). The performance of this new method is evaluated via simulation analysis, showing the good properties of the waveforms in terms of AF performance parameters and in attaining high diversity between waveforms for both fractional and CE fractional libraries. In addition, the applicability of the derived fractional waveforms is experimentally validated, while their performance is evaluated through comparing with conventional techniques in a distributed MIMO radar scenario. Moreover, a novel-multiplexing scheme also based on the FrFT is introduced enabling radar systems to operate in a message exchange mode via embedding the required information into fractional waveforms. The efficiency of the proposed waveform design is evaluated regarding the AF properties of the communicating radar (Co-Radar) waveform. A new, generalised AF is presented based on the KLD and applied in a MIMO radar signal model. The proposed MIMO AF can be factorised into auto-correlation and cross-correlation signal matrices, and channel correlation matrices. Moreover, it is shown that the proposed MIMO AF maximally stretches between 0 and 1, while also being flexible for various geometrical and operating signal configurations. The relationship of the proposed MIMO AF with other definition is also examined, showing that it reduces to the traditional Woodward definition when the same signal model is assumed. In addition, the behaviour of the proposed MIMO AF is investigated for different target placements and operating waveforms highlighting the advantages of each configuration. Finally, the good performance of the AF is demonstrated in a simulated MIMO radar system.
Supervisor: Soraghan, John ; Weiss, Stephan Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral