Pricing-based Distributed Downlink Beamforming in Multi-Cell OFDMA Networks

TL;DR

This paper proposes a distributed beamforming algorithm for multi-cell OFDMA networks that optimizes network utility, converges quickly, and outperforms existing interference mitigation methods, especially under strong interference.

Contribution

It introduces a game-theoretic distributed algorithm with pricing for downlink beamforming, applicable to various utility functions, and analyzes its convergence and efficiency.

Findings

Algorithm converges in few iterations

Significant performance gains under strong interference

Low information exchange overhead

Abstract

We address the problem of downlink beamforming for mitigating the co-channel interference in multi-cell OFDMA networks. Based on the network utility maximization framework, we formulate the problem as a non-convex optimization problem subject to the per-cell power constraints, in which a general utility function of SINR is used to characterize the network performance. Some classical utility functions, such as the proportional fairness utility, the weighted sum-rate utility and the {$\alpha$}-fairness utility, are subsumed as special cases of our formulation. To solve the problem in a distributed fashion, we devise an algorithm based on the non-cooperative game with pricing mechanism. We give a sufficient condition for the convergence of the algorithm to the Nash equilibrium (NE), and analyze the information exchange overhead among the base stations. Moreover, to speed up the optimization of the beam-vectors at each cell, we derive an efficient algorithm to solve for the KKT conditions at each cell. We provide extensive simulation results to demonstrate that the proposed distributed multi-cell beamforming algorithm converges to an NE point in just a few iterations with low information exchange overhead. Moreover, it provides significant performance gains, especially under the strong interference scenario, in comparison with several existing multi-cell interference mitigation schemes, such as the distributed interference alignment method.