Competitive Scheduling for OFDMA Systems with Guaranteed Transmission Rate

Next-generation broadband wireless standards use Orthogonal Frequency Division Multiple Access (OFDMA) as the preferred physical layer multiple access scheme, especially for the downlink. Due to limited resources available at the base station, multi- user resource allocation is crucial for delivering the applications with various quality-of-service (QoS) demands. 

We study the OFDMA resource allocation problem for Adaptive Modulation and Coding (AMC) mode of mobile WiMAX, in which resources in the form of Allocation Units (AU) are to be allocated and the channel knowledge of only some good sub-channels for the users is available to the scheduler. The objective is to maximize the sum-rate for all the users while maintaining guaranteed minimum and maximum traffic rates for certain users under a total transmission power constraint. 

Two novel factors, Band Selectivity Factor (BSF) and Competition Factor (CF), are introduced. Based on the two factors, we propose a sub-optimal approach, Rate-guaranteed Competitive Scheduling (RCS). It is suitable to be deployed in practical systems and exploits multi-user diversity with very low complexity. Simulation results show that RCS can achieve almost 75% of throughput of the optimal solution.