Diversity-Multiplexing Trade-off for Coordinated Direct and Relay Schemes

The recent years have brought a significant body of research on wireless Two-Way Relaying (TWR), where the use of network coding brings an evident advantage in terms of data rates. Yet, TWR scenarios represent only a special case and it is of interest to devise similar techniques in more general multi-flow scenarios. Such techniques can leverage on the two principles used in Wireless Network Coding to design throughput-efficient schemes: (1) aggregation of communication flows and (2) embracing and subsequently cancel/mitigate the interference. Using these principles, we investigate Coordinated Direct/Relay (CDR) schemes, which involve two flows, of a direct and a relayed user. In this paper we characterize a CDR scheme by deriving/bounding the Diversity-Multiplexing Trade-off (DMT) function. Two cases are considered. In the first case a transmitter knows the Channel State Information (CSI) of all the links in the network, while in the second case each node knows only CSI of the links towards its neighbors. The results show that the new CDR scheme outperforms the reference scheme in terms of DMT characterization. Several interesting features are identified with respect to the impact of the CSI knowledge to the improvement in diversity or multiplexing brought by the CDR scheme.The recent years have brought a significant body of research on wireless Two-Way Relaying (TWR), where the use of network coding brings an evident advantage in terms of data rates. Yet, TWR scenarios represent only a special case and it is of interest to devise similar techniques in more general multiflow scenarios. Such techniques can leverage on the two principles used in Wireless Network Coding to design throughput-efficient schemes: (1) aggregation of communication flows and (2) embracing and subsequently cancel/mitigate the interference. Using these principles, we investigate Coordinated Direct/Relay (CDR) schemes, which involve two flows, of a direct and a relayed user. In this paper we characterize a CDR scheme by deriving/bounding the Diversity-Multiplexing Trade-off (DMT) function. Two cases are considered. In the first case a transmitter knows the Channel State Information (CSI) of all the links in the network, while in the second case each node knows only CSI of the links towards its neighbors. The results show that the new CDR scheme outperforms the reference scheme in terms of DMT characterization. Several interesting features are identified with respect to the impact of the CSI knowledge to the improvement in diversity or multiplexing brought by the CDR scheme