Exploiting multiuser diversity using multiple feedback thresholds

Vegard Hassel*, Mohamed-Slim Alouini, David Gesbert, Geir E. Øien

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

46 Scopus citations

Abstract

This paper describes a novel scheduling algorithm that takes advantage of multiuser diversity to obtain the maximum system spectral efficiency and uses multiple feedback thresholds to reduce the feedback load to a minimum. In this scenario the relevant users are probed with a set of carrier-to-noise ratio (CNR) thresholds. The users are first probed with the highest threshold. If none of the users are above this threshold the threshold value is sequentially lowered until one or more users are found. A closed-form expression for the average, normalized feedback load (NFL) is found. It can be argued that it is wise to minimize this average NFL to minimize the guard time needed for the feedback process. Consequently, the optimal CNR thresholds which minimize the average NFL are found. We also develop closed-form expressions for the overall capacity using quantized feedback. Plots show that the number of transmitted symbols between feedback queries has great impact on the overall capacity and that one bit feedback is optimal in all cases. The scheduling outage probability has also been analyzed, and the results show that the scheduling outage probability increases dramatically when a scheduling deadline is exceeded.

Original languageEnglish (US)
Pages (from-to)1302-1306
Number of pages5
JournalIEEE Vehicular Technology Conference
Volume61
Issue number2
StatePublished - Oct 17 2005
Event2005 IEEE 61st Vehicular Technology Conference -VTC 2005 - Spring Stockholm: Paving the Path for a Wireless Future - Stockholm, Sweden
Duration: May 30 2005Jun 1 2005

ASJC Scopus subject areas

  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Exploiting multiuser diversity using multiple feedback thresholds'. Together they form a unique fingerprint.

Cite this