In this paper, we consider the problem of transmission of a delay-sensitive and bursty traffic source over a time-varying Nakagami-m fading channel in a cross-layer optimization framework. Aiming at minimizing the packet delay due to queuing at the data link layer, we present power and rate adaptation policies for coded M-QAM modulation schemes, which guarantee a prescribed channel packet error rate constraint. This is also equivalent to minimizing the system packet loss rate. The proposed adaptation policies are derived both for block and correlated channel fading scenarios. To enable the transmission policy design, we use a statistical model to characterize the packet delay derived from a known result of large deviations theory. Considering the error resolution capability of the automatic repeat request protocol, we then provide the appropriate analytical tools to incorporate the effect of packet retransmission in the proposed optimization framework for transmission policy design. The results show that the proposed adaptation policies compared to others adaptive solutions, significantly improve the delay and throughput performance for delay-sensitive bursty traffic over time-varying fading channels.
Read full abstract