An Improved Penalty-SA Based Resource Allocation Algorithm for OFDMA Cellular Systems

In a multi-cell OFDMA system, the total power is minimized by adopting a centralized approach to adjust the subcarriers and power allocation with the interference from other cells into consideration. In this paper, aiming at simplifying the model and reducing the computational complexity, we improved the Penalty-Simulation Annealing (PSA) resource allocation algorithm proposed by [4]. From the theoretical analysis and simulation results, we observe that the new method can dramatically reduce computational complexity compared with [4] without degrading the performance of the whole system. Compared with the multiallocation algorithm in [2], our method can greatly improve the unit power throughput.