KAUST DepartmentElectrical Engineering Program
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Communication Theory Lab
Permanent link to this recordhttp://hdl.handle.net/10754/563018
MetadataShow full item record
AbstractIn this paper, we study the sum-rate maximization algorithms for downlink and uplink orthogonal frequency division multiple access (OFDMA) systems under proportional-rate constraint (PRC) and minimum-rate constraint. We develop a low-complexity weighted channel signal-to-noise ratio (w-SNR)-based ranking scheme for user selection on each subchannel in OFDMA combined with waterfilling (WF) power allocation. Both offline and online optimization algorithms are developed to optimize the SNR weight vector to maximize the sum rate while satisfying several constraints, such as PRC. The offline weight optimization technique relies on the analytical throughput results developed in this paper, and the online weight adaptation method tracks the user rates and meets the PRC using a subgradient search. Furthermore, we introduce a novel SNR operating region test to enhance the multiuser diversity gain and the sum rate. The proposed schemes have a low complexity, which is linear to the numbers of users and subchannels. Simulation results verify the accuracy of the developed analytical rates and fairness formulas, and show that the proposed w-SNR schemes can achieve higher sum rates than several benchmark schemes that provide the PRC with either short-term or long-term fairness. © 2013 IEEE. © 2013 ESO.
SponsorsThis paper was presented in part at the IEEE ICC Conference, Beijing, China, 2008, and in part at the IEEE GlobeCom Conference, Hawaii, 2009. This work was supported in part by Qatar National Research Fund (QNRF) (a member of Qatar Foundation); in part by the Ministry of Knowledge Economy (MKE), Korea, under the ITRC support program supervised by the NIPA (NIPA-2011-(C1090-1111-0005)); and in part by the US National Science Foundation under Grant No. 0708469, No. 0737297, No. 0837677, the Wright Center for Sensor System Engineering.