Quality of service analysis and improvement with cross layer power allocation in multirelay networks

Quality of service (QoS) guarantee is the main performance metric for designing future wireless networks. For real time applications over wireless links, statistical delay QoS guarantee is the most important QoS metrics. The main goal of this paper is studying the effects of multi-decode-and-forward-relaying transmission on delay QoS guarantee and user satisfaction. For this end, we first propose an approach to simplify the analysis of statistical delay QoS guarantee by the use of a curve named QoS analysis curve. The QoS analysis curve of any system is based on its effective capacity (EC) function. So, we derive a closed-form expression for the EC function of a multi-decode-and-forward-relay network to reduce the computational complexity of finding this quantity. After that, for improving the QoS guarantee performance of a multi-decode-and-forward-relaying system, the EC of the system is maximized by the use of power allocation (PA) technique. At first, an adaptive PA (APA) algorithm is proposed in which the source and relays PA policies are derived analytically. Although, the overhead of this algorithm is large, it significantly improves the statistical delay QoS guarantee performance. Then, for the systems that cannot tolerate the required large overhead of APA, a fixed PA scheme is presented, which is based on particle swarm optimization algorithm. There is a tradeoff between providing the delay QoS guarantee and the required overhead, in selecting APA or fixed PA schemes. The numerical results validate our analytical outcomes and evaluate our proposed algorithms. Copyright (C) 2015 John Wiley & Sons, Ltd.