Distributed energy-efficient design for coordinated multicell downlink transmission

This paper studies joint power allocation and beam-forming for energy efficient communication in coordinated multi-cell multi-user downlink systems. The considered energy efficiency maximization problem which takes both dynamic and static power consumption into account is non-convex and hard to tackle. To address it, the optimization problem is first transformed into a parametric subtractive form using the classical fractional programming method. Then, by introducing the concept of the interference temperature used in cognitive radio networks, the parameterized subtractive form optimization problem is further decomposed into a master problem and a set of subproblems. Based on that, we exploit the convex approximation to develop a decentralized multi-cell multi-user algorithm, which is shown to converge and only needs limited information exchange between the coordinated BSs. Numerical results show that the proposed decentralized energy efficiency algorithm outperforms conventional power allocation algorithms and exhibits a performance close to the optimal centralized solution.