Statistical delay control and QoS-driven power allocation over two-hop wireless relay links

The time-varying feature of wireless channels usually makes the hard delay bound for data transmissions unrealistic to guarantee. In contrast, the statistically-bounded delay with a small violation probability has been widely used for delay quality-of-service (QoS) characterization and evaluation. While existing research on the statistical-delay control mainly focused on the single-hop links, in this paper we propose the QoS-driven power-allocation scheme over two-hop wireless relay links to statistically upper-bound the end-to-end delay for the decode-and-forward (DF) relay transmissions. Specifically, by applying the effective capacity and effective bandwidth theories, we first analyze the delay-bound violation probability over the two-hop link with independent service process in each hop. Then, we show that an efficient approach for statistical-delay QoS guarantees is to make the delay distributions of both hops identical, which, however, needs to be obtained through asymmetric resource allocations over the two hops. Motivated by this observation, we formulate and solve an optimization problem aiming at minimizing the total power consumptions to satisfy the specified end-to-end delay-bound violation probability over two-hop relay links. Also conducted is a set of numerical results to show the impacts of the QoS requirement, the traffic load, and the position of the relay node on the power allocation under our proposed scheme.