The focus of this thesis is peak-to-average power ratio (PAPR) reduction in cooperative wireless networks which exploit orthogonal frequency division multiplexing in transmission. To reduce the PAPR clipping is employed at the source node. The first contribution focuses
upon an amplify-and-forward (AF) type network with four relay nodes which exploits distributed closed loop extended orthogonal space frequency block coding to improve end-to-end performance. Oversampling and filtering are used at the source node to reduce out-of-band interference and the iterative amplitude reconstruction decoding technique is used at the destination node to mitigate in-band distortion which is introduced by the clipping process. In addition, by exploiting quantized group feedback and phase rotation at two of the relay nodes, the system achieves full cooperative diversity in addition to array gain.
The second contribution area is outage probability analysis in the context of multi-relay selection in a cooperative AF network with frequency selective fading channels. The gains of time domain multi-path fading channels with L paths are modeled with an Erlang distribution.
General closed form expressions for the lower and upper bounds of outage probability are derived for arbitrary channel length L as a function of end-to-end signal to noise ratio. This analysis is then extended for the case when single relay selection from an arbitrary number of relay nodes M is performed. The spatial and temporal cooperative diversity gain is then analysed. In addition, exact form of outage probability for multi-path channel length L = 2 and selecting the best single relay from an arbitrary number of relay nodes M is obtained. Moreover, selecting a pair of relays when L = 2 or 3 is additionally analysed.
Finally, the third contribution context is outage probability analysis of a cooperative AF network with single and two relay pair selection from M available relay nodes together with clipping at the source node, which is explicitly modelled. MATLAB and Maple software based simulations are employed throughout the thesis to support the analytical results and assess the performance of algorithms and methods.
A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of Loughborough University.