Low computational complexity channelisation algorithm for multi-standard software defined radio receiver

dc.contributor.advisorMyburgh, Hermanus Carel
dc.contributor.emailu27026494@tuks.co.zaen_ZA
dc.contributor.postgraduateOtunniyi, Temidayo Oluwafunke
dc.date.accessioned2022-02-28T13:12:04Z
dc.date.available2022-02-28T13:12:04Z
dc.date.created2022-04
dc.date.issued2022-01
dc.descriptionThesis (PhD (Computer Engineering))--University of Pretoria, 2022.en_ZA
dc.description.abstractThis work focuses on achieving a low complexity channelisation algorithm for these two algorithms for any frequency \change[Friend]{bands}{band} whether low or high, exploring three approaches in developmental sequence based on different computational rationales. The first approach involves the development of two algorithms described as Hybrid Generalised Discrete Fourier Transform (HGDFT) and Hybrid Farrow (HFarrow) Channelisation Algorithms. These developments involve hybridising frequency response masking techniques and coefficient decimating filter with either the modulated GDFT or modulated Farrow filter. These methods efficiently reduce the computational complexity to some extent. The second approach focuses on improving the performance of HGDFT and HFarrow algorithms using different digital number systems, implemented on parallel distributed arithmetic architecture. The digital number representations used are: Parallel Distributed Arithmetic Based Residual Number System (PDA-RNS), Parallel Distributed Arithmetic Based Canonical Signed Residual Number System (PDA-CSRNS), and Parallel Distributed Arithmetic Based Common Sub-expression Elimination Method (PDA-CSE). These approaches resulted in significantly fewer filter coefficients as well as a reduction in the number of multipliers and adders used. Thus, the computational complexities of HGDFT and HFarrow improved remarkably when optimised with the parallel distributed arithmetic and number systems. However, the filter's passband ripples and the stop band attenuation degrade in performances due to the rounding of the continuous filter coefficients. A Genetic Algorithm (GA) was introduced to optimise the performances of the hybrid filter. Results obtained showed better filter structures with lower computational complexities. The third approach uses the multi-rate, multi-stage method to reduce computational capability of the HGDFT and HFarrow channelisation algorithms. The large number of filter coefficients in HGDFT and HFarrow were factored into different stages. These relaxed the filter specifications and resulted in more reduced computational complexities.en_ZA
dc.description.availabilityUnrestricteden_ZA
dc.description.degreePhD (Computer Engineering)en_ZA
dc.description.departmentElectrical, Electronic and Computer Engineeringen_ZA
dc.identifier.citation*en_ZA
dc.identifier.otherA2022en_ZA
dc.identifier.urihttp://hdl.handle.net/2263/84271
dc.language.isoenen_ZA
dc.publisherUniversity of Pretoria
dc.rights© 2022 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria.
dc.subjectUCTD
dc.subjectComputational complexity
dc.subjectCommon subexpression
dc.subjectCanonical signed digit
dc.subjectCommon subexpression elimination method
dc.subjectParallel distributed algorithm
dc.titleLow computational complexity channelisation algorithm for multi-standard software defined radio receiveren_ZA
dc.typeThesisen_ZA

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Otunniyi_Computational_2022.pdf
Size:
9.31 MB
Format:
Adobe Portable Document Format
Description:
Thesis

License bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
1.75 KB
Format:
Item-specific license agreed upon to submission
Description: