Distributed resource allocation for single-hop networks under the SINR model

We consider the problem of distributed resource allocation in a single-hop wireless network under the SINR model. The network consists of fixed transmitters and their distinct receivers, and the physical layer model we consider is the SINR threshold model, where the transmitter will be able to send data successfully to its receiver at a fixed rate if the SINR at the receiver is greater than a given threshold. We address the problem of joint power control and link activation in a distributed setting, in order to stably support any traffic arrival process whose rate vector lies inside the capacity region. One of the technical contributions of this paper is in identifying a linear program to determine the optimal durations for employing each activation, whose solution decouples into a distributed algorithm, upon introducing a logarithmic barrier function.