A convex optimization approach to decentralized rate control in wireless networks with partial interference

Because wireless networks use shared communication channels, contention and interference can significantly degrade throughput and fairness. Optimal rate control algorithms can be designed for wireless networks by modeling the contention between routers using capacity constraints and solving a convex optimization problem. In this work we develop a more accurate network model that directly incorporates partial interference as a receiving constraint, so that it is modeled separately from contention. We show that using this new model leads to a convex optimization problem when formulated using link rates, but it is non-convex when formulated using flow rates.We then use numerical results to illustrate situations when modeling partial interference separately yields significantly higher effective utilities in practical network topologies.