Uplink massive MIMO systems under statistical-queueing constraints

Cross-layer design is a suitable approach that can address some of throughput challenges for future multimedia applications. In this regard, effective capacity concept offers a suitable metric to assess the implications that physical layer design may have on link layer performance. However, most reported studies in this aspect consider the case of downlink scenario for convenience multiple input multiple output (MIMO) systems. This paper analyses the uplink-effective capacity in a multi user massive MIMO system considering the quality of service (QoS) statistical-exponent constraints. A location-dependent and a location-independent (averaged out) closed form expressions are derived for the effective capacities in asymptotically large receive antenna system with perfect and imperfect channel state information (CSI) scenarios. In addition, expressions for the asymptotically high signal-to-noise ratio regimes are provided. Obtained results show that the effective capacity performance is a monotonically increasing function of the base station receive antennas (N_r) and the large scale shadowing parameter. Moreover, increasing N_r by almost 74 antennas for imperfect CSI, can compensate for system capacity degradation due to queueing delay constraint of θ = 1.9×10^-4 per bit.