Network coding for the multiple access relay channel

To cope with the growing demands of wireless applications and mobile internet, wireless communications is expected to provide high data rate. Cooperative communications has been presented to increase the throughput, the reliability and the robustness of the wireless system. Recently, network coding has been introduced to further increase the throughput. In contrast to the conventional network coding techniques where they are performed in the upper layer of the protocol stack, in this thesis we focus on network coding techniques which are performed in the physical layer for the multiaccess relay channel (MARC) system. In the first part of the thesis, we consider a two-user multiaccess relay channel where the relay implements decode-and-forward strategy. In this model, the users are allowed to simultaneously transmit their messages on the same channel. This is a form of physical-layer network coding where simultaneous transmissions are observed by the relay as a linear combination. For this transmission scheme, we discuss the design criteria and evaluate the achievable sum-rate. Also, we study the problem of allocating the resources optimally in order to maximize the achievable sum-rate. We show that by allowing the users to simultaneously transmit their messages on the same channel, one can afford a larger sum-rate compared to the case where the users transmit their messages on orthogonal channels. In the second part of the thesis, we further study the MARC model. Based on the compute-and-forward scheme, we establish three coding schemes that are of network coding spirit. In these schemes, the destination does not decode the information messages directly from its output, but uses it to recover two linearly independent integer-valued combinations that relate the transmitted messages. The three schemes differ essentially through the operations implemented by the relay. In the first one, the relay implements compute-and-forward. In the second one, the relay implements compress-and-forward and in the third one the relay implements amplify-and-forward. We show that the first scheme can outperform standard relaying techniques in certain regimes. Also, we show that the second and third schemes, while relying on feasible structured lattice codes, can at best achieve the same performance as regular compress-and-forward and amplify-and-forward, respectively. In the last part of the thesis, we consider the multi-user multi-relay network with compute-and-forward strategy. We propose an efficient algorithm to allocate the powers at the users and to find the integer coefficients of the linear combinations in such a way that the transmission rate is maximized and that the computed linear combinations are linearly independent.