Performance of Practical Subcarrier Allocation Schemes for OFDMA

In this paper, we consider the problem of adaptive resource allocation for OFDMA. Aiming at practical (low-complexity) algorithms we study disjoint subcarrier and power allocation. A number of low-complexity subcarrier allocation strategies are discussed. We study "user by user" subcarrier assignment (i.e. the iteration over users) based on a user ordering according to the user average power gains as well as "subcarrier by subcarrier" assignment (i.e. the iteration over subcarriers) employing subcarrier ordering according to different metrics. By employing the order statistics theory we prove that the iteration over users starting from the worst user outperforms other orderings in Rayleigh environment. Efficiency of strengthening the considered subcarrier allocation algorithms by a non-iterative power loading technique is also studied.