Node-Based Distributed Optimal Control of Wireless Networks

We present a unified analytical framework within which power control, routing, and congestion control for wireless networks can be optimized on a node-by-node basis. We consider a multi-commodity flow model for interference-limited wireless networks, and develop distributed scaled gradient projection algorithms which iteratively adjust power control and routing schemes at individual nodes to minimize convex link costs. We provide locally computable scaling matrices which guarantee fast convergence of the algorithms to the global optimum from any initial condition. Furthermore, we show that congestion control can be seamlessly incorporated into our framework with the introduction of virtual overflow links.