Distributed throughput-maximization using the up- and downlink duality in wireless networks

We consider the throughput-maximization problem for both up- and downlink in the wireless network. For the purpose, we design an iterative and distributive uplink algorithm based on Lagrangian relaxation. Using the Lagrange multipliers and the network duality, we make throughput-maximization in the downlink. Our analysis and computational tests assume that channels are symmetric between up- and downlink. The network duality we proved here is a generalized version of previous researches by Jindal et al. and Catrein et al. Computational test shows that the performance of up- and downlink throughput for our algorithms is close to the optimal value for the channel orthogonality factor between 0.5 and 1. In particular, our duality-based approach gives 97-98% throughput of the optimal uplink algorithm proposed by Kumaran and Quian, and a downlink heuristic algorithm (MPA-1) proposed by Mo and Kim, when the channel orthogonality factor is a value between 0.5 and 1. On the other hand, when channels are rather orthogonal (between 0 and 0.5), we have observed some throughput degradation in the downlink, which is about 86% of MPA-1. Considering the complexity of the existing algorithms, we conclude that these results are quite encouraging in terms of both performance and practical applicability of the generalized duality theorem.