On the Adaptive Sum-Capacity of Distributed Multiple Access with Individual CSI

In wireless communications, a fading multiple-access channel (MAC) is typically used to model the uplink communication. Conventional MACs assume a centralized system, where the transmission rate and power are chosen centrally for every fading vector realization. On the other hand, there is considerable interest in the performance of distributed multiple access systems, where the lack of global channel state information (CSI) demands novel communication strategies. We consider a block-fading MAC where each transmitter is aware only of its own link CSI, which we term as the individual CSI MAC. The receiver has access to the full CSI of all links. This model was recently introduced in the information theory literature, and naturally leads to a distributed access system with several applications. An important utility of interest for this model is known as the power-controlled adaptive sum-capacity, whose evaluation is an open problem. This is the main subject of the current paper. We present the power-controlled adaptive sum-capacity of a wide class of popular fading MAC models. In particular, we characterize the sum-capacity when the statistics of the channel are identical across users. The proposed schemes also allow a low-complexity successive-cancellation decoding using rate-splitting. Furthermore, the optimal schemes are extended to situations in which each transmitter has additional finite-rate partial CSI on the link quality of others.