Optimal cooperative water-filling power allocation for OFDM system

Hui Wang,Qimei Cui,Xiaofeng Tao,Mikko Valkama,Y. Jay Guo
DOI: https://doi.org/10.1109/WCNC.2013.6555170
2013-01-01
Abstract:It is well known that traditional water-filling provides a closed form solution for capacity maximization in orthogonal frequency division multiplex (OFDM) system. In this paper, cooperative power allocation is investigated in a two-transmitter multi-receiver model for OFDM systems. The local full channel state information (CSI) is available at the two transmitters respectively, where each transmitter has an individual power constrain. The transmitters first cooperate by sharing CSI, and then jointly optimize power allocation in the metric of sum throughput, which can be modeled as a convex optimization problem. Through an application of Karush-KuhnTucker (KKT) conditions, the convex optimization problem is reformulated as a simplified convex one. Then the closed form solution is derived, which takes a form similar to classic waterfilling principle. Based on the solution, the optimal cooperative power allocation algorithm is constructed, the structure of which can be explained as a cooperative water-filling relative to the traditional water-filling. Finally, numerical simulation is given to evaluate and demonstrate the performance of the optimal cooperative water-filling scheme.
What problem does this paper attempt to address?