Adaptive scheduling with postexamining user selection under nonidentical fading
KAUST DepartmentCommunication Theory Lab
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Electrical Engineering Program
Physical Science and Engineering (PSE) Division
Permanent link to this recordhttp://hdl.handle.net/10754/562393
MetadataShow full item record
AbstractThis paper investigates an adaptive scheduling algorithm for multiuser environments with statistically independent but nonidentically distributed (i.n.d.) channel conditions. The algorithm aims to reduce feedback load by sequentially and arbitrarily examining the user channels. It also provides improved performance by realizing postexamining best user selection. The first part of the paper presents new formulations for the statistics of the signal-to-noise ratio (SNR) of the scheduled user under i.n.d. channel conditions. The second part capitalizes on the findings in the first part and presents various performance and processing complexity measures for adaptive discrete-time transmission. The results are then extended to investigate the effect of outdated channel estimates on the statistics of the scheduled user SNR, as well as some performance measures. Numerical results are provided to clarify the usefulness of the scheduling algorithm under perfect or outdated channel estimates. © 1967-2012 IEEE.