A distributed solution for proportional fairness optimization in load coupled OFDMA networks

We propose a novel distributed downlink technique for proportional fairness optimization in load coupled orthogonal frequency multiple access (OFDMA) networks where the power allocation, the rate allocation and resources assignment are jointly optimized. Building upon recent results on nonlinear interference models for OFDMA-based wireless systems, we pose the problem of finding proportionally fair rates as a convex network utility optimization problem for which solutions can be computed efficiently and are known to be energy efficient. Furthermore, we propose a distributed optimality-achieving implementation based on a message passing protocol between base stations. Simulations show a potential for improvements in terms of network utility and energy consumption.