Centralized power allocation algorithms for OFDM cellular networks

The throughput maximization problem of a multicell orthogonal frequency division multiplexing (OFDM) network is studied by considering power allocation and adaptive modulation techniques. The throughput maximization problem is formulated as a nonlinear programming problem and the optimum power vector is determined by using the feasible sequential quadratic programming (FSQP). The disadvantage of the FSQP is its slow convergence especially when the number of optimized variables is large. To overcome this drawback, we propose a low-complexity suboptimal power allocation algorithm which updates the power vectors iteratively based on the first order necessary conditions for local optimum points. Simulation results show the proposed power allocation algorithms increase system throughputs from 5% to 30% over the equal power allocation scheme in various channel conditions.