Energy efficiency in multiple-antenna channels with markov arrivals and queueing constraints

Energy efficiency in multiple-antenna fading channels is analyzed in the presence of Markov sources and queueing constraints, which are imposed as limitations on buffer overflow probabilities. Two random arrival models, namely discrete Markov and Markov fluid processes, are considered. Employing the notions effective capacity of time-varying channels and effective bandwidth of time-varying sources, maximum average arrival rates of these sources that can be supported by multiple-antenna wireless systems under statistical queueing constraints are determined and the throughput levels are identified. In the low signal-to-noise ratio (SNR) regime, minimum energy per bit and wideband slope expressions are obtained. Performance with both uniform power allocation and low-SNR optimal power allocation across transmit antennas is investigated. It is shown that the minimum energy per bit does not depend on the queueing constraints and source burstiness. On the other hand, wideband slope is shown to decrease as queueing constraints get stricter and/or sources become more bursty, thus resulting in degraded energy efficiency.