LAUSR

In this paper, we investigate energy-efficient downlink resource allocation in heterogeneous orthogonal frequency division multiple access (OFDMA) networks, and formulate the energy efficiency (EE) maximization problem as a mixed-integer nonlinear fractional programing (MINLFP) problem with a nonconcave nonlinear objective function and nonlinear constraints. By means of fractional programing and changing of variables, we transform the original MINLFP problem into an equivalent optimization problem in a parametric subtractive form, which is proved to be a concave mixed-integer nonlinear programing (MINLP) problem and is optimally solved by using Dinkelbach and branch-and-bound (BB) methods. In BB method, the concave MINLP problem is relaxed to a series of concave nonlinear programing problems and solved by the use of Powell–Hestenes–Rockafellar augmented Lagrangian (PHR-AL) method. The optimal solution can be used to benchmark the performance of suboptimal solutions. As the computational complexity of BB method increases exponentially with problem size, we further develop a suboptimal two-step scheme, which first allocates the resource blocks and then performs the transmit power control to give suboptimal solution with much lower complexity. Simulation results demonstrate the effectiveness of the proposed schemes and show that the proposed suboptimal two-step scheme is promising for practical applications as it makes a good tradeoff between EE performance and computational complexity.