Achievable Rate of Two-Hop Channels under Statistical Delay Constraints.

Deli Qiao
DOI: https://doi.org/10.1109/glocom.2015.7417331
2015-01-01
Abstract:In this work, the performance of two-hop wireless systems under statistical queueing \emph{end-to-end} delay constraints is investigated. In the two-hop links, it is assumed that there is no direct link between the source and destination. The communication between the source and destination is accomplished via an intermediate full-duplex relay node, which forwards the information to the destination by employing the decode-and-forward scheme. Both the queues at the source and relay node are subject to statistical queueing constraints imposed on the limitations of buffer violation probability. Given statistical delay constraints specified via maximum delay and delay violation probability, the tradeoff between the statistical delay constraints imposed to any two concatenated queues is identified. With this characterization, the maximum constant arrival rates that can be supported by this two-hop link are obtained by determining the effective capacity of such links as a function of the statistical delay constraints and signal-to-noise ratios (SNR) at the source and relay, and the fading distributions of the links. It is shown that the tradeoff between statistical delay constraints of the queues at the source and relay can improve the achievable rate. Also, compared with the performance achieved without buffer at the relay, the throughput with buffer at the relay can be larger for certain cases. Overall, the impact of statistical delay constraints on the achievable throughput is provided.
What problem does this paper attempt to address?