OFDMA with Resource and Traffic Constraints: Sum Rate Maximization with no CSI

For a specific high-throughput wireless local area network based on OFDM we study the sum rate of time and frequency division multiple access (TDMA, OFDMA). In allocating resources we consider a minimum size of the smallest resource unit that a user can be allotted. We further assume packet switched traffic with most of it carried by only a few large packets. Relatively little data is carried by a large number of small packets. We show that given the resource and traffic constraints the sum rate may be drastically reduced compared to the unconstrained case in TDMA. We also show that by allowing many users to be active simultaneously OFDMA maintains high throughput. In OFDMA we can do time-sharing between two different bandwidth allocations. We derive the number and rates of users in each of the modes to maximize the sum rate.