Sequential Geometric Programming for 2 × 2 Interference Channel Power Control

Abstract

We analyze the performance of sequential geometric programming (SGP) in solving the nonconvex power control problem of maximizing the sum capacity of the interference channel with no multiuser decoding. We focus on the 2x2 interference channel, for which we know the optimal power allocation tuples. We validate the SGP approaches of single and double condensation methods and we analyze the accuracy and convergence performance of each.