Queuing Analysis on Adaptive Transmission in MIMO Systems with Imperfect CSI

The adaptive transmission jointly considering adaptive modulation and coding (AMC) and automatic repeat request (ARQ) has been applied to MIMO systems to provide high spectral efficiency and link robustness. However, the adaptive transmission requires knowledge of the channel state information (CSI) and in practice perfect CSI is rarely available. In this paper, under the condition of finite-length buffer, the queuing characteristics of the Zero-Forcing (ZF) detection MIMO systems combining with adaptive transmission in the presence of imperfect CSI are analyzed. We use Markov-chain-based analytical framework to investigate the performance of the systems and derive the close form expressions for the throughput, packet loss rate and average delay, from which we can maximize the overall system throughput under the specified QoS constraints through numerical search. The simulation results demonstrate the verification of the performance analysis.