Dynamic subchannel and slot allocation for OFDMA networks supporting mixed traffic: upper bound and a heuristic algorithm

Optimizing resource allocation over only the frequency dimension in a single-cell OFDMA network is strictly suboptimal when heterogeneous non-real-time data users coexist. In this work we examine how to jointly exploit multi-user diversity on frequency and time domains towards maximizing cell throughput performance and satisfying individual rate and fairness QoS constraints. We develop a performance upper bound as well as an efficient heuristic algorithm for subcarrier and slot allocation which outperforms existing algorithms under polynomial complexity.