On the asymptotic throughput of OFDMA systems with best-M feedback

On the asymptotic throughput of downlink OFDMA systems with best-M CQI feedback
Nicola Varanese
PhD Student

DEI - Aula 3B
10 Novembre 2010
Ore 11.00


Orthogonal frequency-division multiple access (OFDMA) is employed by wideband cellular communication systems to cope with the deleterious effects of multipath propagation. User scheduling in the frequency domain requires each user to feed back channel quality information (CQI) relative to each transmission resource (sub-band) to the base station. CQI feedback entails a heavy overhead on the reverse channel and calls for the employment of efficient compression techniques. 3GPP E-UTRA (LTE) systems employ the so-called “best-M” compression scheme, which prescribes each user to report the index of the M best sub-bands along with the rate supported on those resources. In this work, the performance of OFDMA user scheduling with best-M CQI feedback is investigated when the number of users grows large. In particular, we derive a useful analytical approximation of the system throughput for both single and multi-antenna communication schemes. A channel hardening effect is observed, whereby channel variability and multiuser diversity gains reduce for increasing M . We show by simulations that the approximation is particularly tight for small M and in the single antenna case. Even when the approximation is not tight, it is still insightful since it allows to infer how the system throughput scales with the number of users.

Nicola Varanese

Area di ricerca: