Achieving Full Diversity by Selection in Arbitrary Multi-Hop Amplify-and-Forward Relay Networks

Although several cooperative diversity strategies have been proposed for simple 2-hop multiple-relay networks, practical and efficient strategies that guarantee maximum diversity order for arbitrary, multi-hop amplify-and-forward relay networks are not yet available. In this paper, the maximum achievable diversity order for a given connectivity is shown to be determined by the number of links on the minimum cut set, and achieved by a selection strategy which maximizes end-to-end SNR. The optimal selection requires the channel information of all the links and serves as a performance bound. A sub-optimal selection method is then proposed to achieve full diversity in a distributed and low-complexity manner. The performance loss of the sub-optimal selection is shown to be small by both theoretic analysis and simulations.