Maximizing throughput in cooperative networks via cross-layer adaptive designs

For a three node cooperative wireless network with an overall transmission power constraint, we present an adaptive cross-layer design framework for maximization of a single user throughput using the power allocation ratio among the source and relay nodes, the packet length, and the modulation level as optimization variables. The design framework integrates automatic repeat request (ARQ) at the MAC layer and cooperative diversity at the physical (PHY) layer. We consider amplify-and-forward (AF) and selective decode-and-forward (SDF) schemes in our study and investigate the impact of average channel qualities on throughput maximization. We show that at low SNR regime, the maximum throughput is achieved by adapting the power allocation between the source and the relay nodes, whereas at high SNR the maximum throughput is attained by adapting all three parameters. While the two cooperative schemes have the same diversity gain, we show that the SDF scheme provides a slightly higher throughput than AF.