This paper proposes an optimal architecture for wireless networks based on layers and layer interfaces. In the presence of fading the architecture is shown to be optimal. The result follows from a subgradient descent algorithm on the dual function of a generic wireless networking optimization problem. The fact that these non-convex optimization problems have nonetheless zero duality gap is exploited.
Date of this Version
concave programming, duality (mathematics), gradient methods, network interfaces, radio networks, layers-and-layer interface, nonconvex optimization problem, subgradient descent algorithm, wireless network optimization problem, zero duality gap, Cross-layer design, Optimization, Wireless networks
Date Posted: 13 October 2009