Relay-assisted communication has been shown to be an effective technique to improve the reliability and throughput of real-world wireless communication networks. It enables single-antenna users to form a virtual antenna array without installing multiple antennas at the transmitter or the receiver. As one of the well-known relaying protocols, the decode-and-forward (DF) protocol is widely used and studied. In this thesis, we re-examine the decoding problem of the coded DF relay channel. The error-free decoding process at the relay is typically assumed by the conventional maximal-ratio combining (MRC) based decoder at the destination. However, one of the important and practical issues of the DF protocol is the unsuccessful decoding at the relay, which can cause error propagation especiall...[ Read more ]

Relay-assisted communication has been shown to be an effective technique to improve the reliability and throughput of real-world wireless communication networks. It enables single-antenna users to form a virtual antenna array without installing multiple antennas at the transmitter or the receiver. As one of the well-known relaying protocols, the decode-and-forward (DF) protocol is widely used and studied. In this thesis, we re-examine the decoding problem of the coded DF relay channel. The error-free decoding process at the relay is typically assumed by the conventional maximal-ratio combining (MRC) based decoder at the destination. However, one of the important and practical issues of the DF protocol is the unsuccessful decoding at the relay, which can cause error propagation especially when the quality of the source-to-relay link is poor.

In this thesis, we first re-examine the decoding problem in the convolutionally coded relay channel with the classical decode-and-forward protocol. By tackling the challenge of modeling the error propagation effect at the relay, a near BER-optimal decoding (NBOD) algorithm at the destination is derived, assuming the availability of perfect receiver channel state information. Its decoding complexity is linear in the information block length, while the exact BER-optimal decoding algorithm with a sub-exponential complexity is still unknown. The proposed NBOD algorithm can perform close to the near maximum likelihood decoding (MLD) performance bound on BER in the three-node one-way relay channel scenario. In addition, we further extend the proposed algorithm for more general single-source single-destination DF relay networks.

We further study the decoding problem for a distributed Turbo code (DTC). We first propose the corresponding trellis error model in this scenario to tackle the challenge of representing all the possible relay decoding errors and characterizing the statistics of relay decoding errors accurately under the BPSK modulation. The only major approximation involved in the derivation of the trellis error model is the pairwise error probability approximation. Accordingly, two trellis-error-model-based iterative decoders (TEM-ID) with two and three components are derived, respectively. By adding one more component, the TEM-ID with three components can reduce the complexity without incurring much performance loss. Moreover, we propose a near MLD performance bound on BER of the DTC to evaluate the derived decoders. The extrinsic information transfer (EXIT) chart analysis is then performed to characterize the asymptotic gain achieved by our proposed decoders for large code lengths. In addition, we extend the proposed decoders to general distributed concatenated coding systems with DF-based relays.

Besides, we study the decoding problem for the LDPC codes over relay channels. We propose a novel Tanner graph error model to represent the relay decoding errors and characterize their statistics. Accordingly, a Tanner-graph-error-model-based message passing (TGEM-MP) decoder is developed at the destination, which is implemented on an extended Tanner graph. Furthermore, the proposed error model can transfer the relay channel as a traditional point-to-point channel in a more accurate way, which allows us to apply the techniques used in the point-to-point channel to the relay channels. In this thesis, we perform the density evolution (DE) to investigate the convergence property of the proposed decoder. Besides, the proposed decoder can be applied to the convolutionally coded relay channels to realize the near BER-optimal decoding algorithm such that the complexity is independent of the number of states of the underlying convolutional code. Moreover, our decoder can be extended to the bit-interleaved coded modulation system with any high order modulation schemes.

Finally, we re-examine the decoding problem of the general two-way relay channel with the application of binary network coding. The proposed decoders based on the trellis error model and Tanner graph error model can be used in this scenario. We consider the practical convolutional codes and LDPC codes, which are included in the wireless Local Area Network (LAN) standard (i.e., IEEE802.11n) with different modulation schemes. Simulation results show that our decoders can offer significant performance improvement over the conventional decoders.