Generalized model for scheduling in MIMO multiple access systems: A cross-layer approach

Research on scheduling in MIMO multiple access systems is traditionally carried out from an information-theoretic point of view. Such information-theoretic approaches are feasible in homogeneous systems where, although channel fading is assumed, all terminals use the same modulation format, channel coding scheme, data transmission rate, packet lengths and no traffic issues are considered. In these systems, to select the set of terminals that maximize the instantaneous sum capacity leads to the highest system throughput. Unfortunately, in heterogeneous systems this assumption is not always true. Consequently, there is a need to evaluate the effects of such heterogeneity from a network point of view. In this paper, following a cross-layer approach, a generalized model for scheduling in MIMO multiple access is presented. Such model allows to evaluate the effects of heterogeneity on system throughput. Particularly, this generalized model accommodates any kind of heterogeneity from multiuser diversity due to channel fading to heterogeneity due to different data transmission rates. Different scheduling policies are analyzed under this generalized model and further applied to a sensor network. The results obtained are useful to achieve practical insight of such scheduling policies.