Fixed Versus Selective Scheduling for Buffer-Aided Diamond Relay Systems Under Statistical Delay Constraints

In this paper, the achievable rate of buffer-aided diamond relay systems under delay constraints is studied. The effective capacity, characterizing the maximum constant arrival rate to the source while satisfying the statistical end-to-end delay constraints, is obtained for the conventional relay selection protocol, fixed decode-and-forward relaying schemes, and selective forwarding protocols. A selection forwarding policy that maximizes the delay exponent at the source is proposed, which is shown to be independent of delay constraints and select the relay with better channel conditions to forward the messages. Through numerical results, it is demonstrated that the selection forwarding strategy achieves superior performance than the conventional relay selection and fixed scheduling schemes under a variety of delay constraints.