Optimal Subcarrier Sharing for Weighted Sum of Rates Maximization in Multiuser-OFDM Systems

We consider the maximization of the weighted sum of rates in a multiuser OFDM system with dynamic subcarrier allocation and subcarrier-wise adaptive modulation and power. We assume a total peak power constraint and a target bit error rate (BER). We allow each subcarrier to be shared by more than one user. This sharing implies an interference which the receivers deal with as noise. In this paper, we show that the resulting non-convex optimization problem can be decomposed into two tractable subproblems: a subcarrier assignment and a power allocation. We prove that the optimal subcarrier assignment is exclusive, i.e. each subcarrier has to be allocated to one user only. The optimal power allocation corresponds to a multilevel water-filling. Since the solutions to these two subproblems are mutually dependent, one can use exhaustive search to explicit the optimal subcarrier assignment. To avoid the burden of exhaustive search, we show that a close-to-optimal low-complexity solution is obtained if the total power is equally partitioned among the subcarriers. Finally, simulation results show that the proposed suboptimal solution yields a significant gain in terms of average weighted sum of rates compared to a traditional static subcarrier allocation.