Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.543861
Title: Strategies for devising automatic signal recognition algorithms in a shared radio environment
Author: Wagstaff, Adrian John
ISNI:       0000 0004 2708 9172
Awarding Body: Open University
Current Institution: Open University
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
In an increasingly congested and complex radio environment interference is to be expected, which poses problems for Automatic Signal Recognition (ASR) systems. This thesis explores strategies for improving ASR performance in the presence of interference. The thesis breaks the overall research question down into a number of subquestions and explores each of these in turn. A Phase-symmetric Cross Recurrence Plot is developed and used to show how a radio signal can be manipulated to separate information about the modulation from the information being carried. The Logarithmic Cyclic frequency Domain Profile is introduced to illustrate how a logarithmic representation can be used for analysing mixtures of signals with very different cyclic frequencies. After defining a canonical ASR system architecture, the concepts of an Ideal Feature and Interference Selectivity are introduced and applied to typical features used in ASR processing. Finally it is shown how these algorithmic developments can be combined in a Bayesian chain implementation that can accommodate a wide variety of feature extraction algorithms. It is concluded that future ASR systems will require features that can handle a wide range of signal types with much higher levels of interference selectivity if they are to achieve acceptable performance in shared spectrum bands. Intelligent segmentation is shown to be a requirement for future ASR systems unless features can be developed that have near ideal performance.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.543861  DOI:
Share: