Abstract
Cooperative relaying communication has attracted increasing attentions for distributed spatial diversity gain via resource sharing between mobile terminals (MTs), without implementation of multiple antennas on size-limited MTs. Motivated by high data rate transmission in communication systems, relaying techniques need to be combined with advanced air interface such as orthogonal frequency division multiplexing (OFDM). In this thesis, we investigate OFDM-based relaying communications in terrestrial and satellite systems. We first analyze bit error rate (BER) performance of OFDM-based selective decode and forward (SDF) relaying with bit interleaved coded modulation (BICM) coding scheme, and obtain a sub-optimal signal-to-noise ratio (SNR) threshold via optimization of derived BER upper bound. This proposed sub-optimal SNR threshold scheme not only shows significant cooperative diversity in comparison to the single link, but also has almost the same performance to the scheme with optimal SNR threshold obtained by exhaustive search. Secondly, in satellite broadcast networks, a spectrally efficient SDF relaying architecture with BICM-OFDM air interface is proposed to improve system performance by exploiting cooperative diversity. For efficient use of frequency resources, the terrestrial relaying retransmission share the same frequency band with the satellite broadcasting, which causes cochannel interference (CCI). We then employ iterative signal processing techniques, such as expectation maximization (EM) channel estimation, parallel interference cancellation (PIC), Bayesian detector, and maximum a posteriori (MAP) decoder, to cancel CCI and achieve the cooperative diversity. In computer simulations, the proposed spectrally efficient SDF relaying scheme demonstrates significant cooperative diversity and close performance to the benchmark scheme with perfect channel state information (CSI). Thirdly, we investigate the cooperative diversity for uncoded OFDM-based amplify and forward (AF) relaying with residual carrier frequency offsets (CFOs). By using Characteristic function (CHF) and Beaulieu series, a conditional BER model is achieved. Then numerical Monte-Carlo techniques are exploited to average the conditional BER expression over CSI and CFO distributions. This semi-analytical BER model shows high accuracy in comparison to simulation results. Moreover, the OFDM-based AF relaying shows obvious cooperative diversity gain for residual CFO with small variances. Last but not the least, to improve power efficiency, we perform bit and power loading (BPL) for OFDM-based two-way AF relaying. For the case nodes don’t have their local power constraints, we derive the flexible power ratio allocation via optimization of received SNRs. On the contrary, with local power constraints, we fix the transmitted power ratio between the relay and two sources, due to high optimization complexity. Then greedy algorithm is employed to perform BPL in an iterative way. Our proposed BPL scheme shows large performance gain in symbol error rate (SER) compared with nonadaptive transmissions.