Two key technology enablers for next generation networks are examined in this thesis, namely Cognitive Radio (CR) and Spectrally Efficient Frequency Division Multiplexing (SEFDM). The first part proposes the use of traffic prediction in CR systems to improve the Quality of Service (QoS) for CR users. A framework is presented which allows CR users to capture a frequency slot in an idle licensed channel occupied by primary users. This is achieved by using CR to sense and select target spectrum bands combined with traffic prediction to determine the optimum channel-sensing order. The latter part of this thesis considers the design, practical implementation and performance evaluation of SEFDM. The key challenge that arises in SEFDM is the self-created interference which complicates the design of receiver architectures. Previous work has focused on the development of sophisticated detection algorithms, however, these suffer from an impractical computational complexity. Consequently, the aim of this work is two-fold; first, to reduce the complexity of existing algorithms to make them better-suited for application in the real world; second, to develop hardware prototypes to assess the feasibility of employing SEFDM in practical systems. The impact of oversampling and fixed-point effects on the performance of SEFDM is initially determined, followed by the design and implementation of linear detection techniques using Field Programmable Gate Arrays (FPGAs). The performance of these FPGA based linear receivers is evaluated in terms of throughput, resource utilisation and Bit Error Rate (BER). Finally, variants of the Sphere Decoding (SD) algorithm are investigated to ameliorate the error performance of SEFDM systems with targeted reduction in complexity. The Fixed SD (FSD) algorithm is implemented on a Digital Signal Processor (DSP) to measure its computational complexity. Modified sorting and decomposition strategies are then applied to this FSD algorithm offering trade-offs between execution speed and BER.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:587673 |
| Date | January 2013 |
| Creators | Grammenos, R. C. |
| Publisher | University College London (University of London) |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://discovery.ucl.ac.uk/1394409/ |
Page generated in 0.0037 seconds