Skip to main navigation Skip to search Skip to main content

Downlink beamforming in multi-antenna two-tier networks with user selection

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

We consider a two-tier cellular network with universal frequency reuse, consisting of a macrocell overlaid with shorter range, low power multi-antenna base stations. Each transmitter selects a set of users among all active users in its cell and serves them using multiuser linear preceding. The effect of user selection in throughput and spatial reuse is investigated and analytical expressions and bounds for the success probability, area spectral efficiency, and average user rate are derived. Our results indicate that downlink multiuser transmission shows throughput gains at both tiers as compared to single-user communications, if combined with scheduling, while the optimal number of served users per cell depends on different network parameters such as pathloss exponent, interférer density, and outage constraints. A multi-mode switching strategy, which instantaneously adapts the number of served users, is then proposed. Furthermore, we provide insights on efficient user selection metrics. Both analytical and numerical results show that zero-forcing preceding may increase coverage and spatial reuse by reducing both intra-tier and cross-tier interference through user scheduling.

Original languageEnglish
Title of host publication2011 IEEE GLOBECOM Workshops, GC Wkshps 2011
Pages258-263
Number of pages6
DOIs
StatePublished - 2011
Externally publishedYes
Event2011 IEEE GLOBECOM Workshops, GC Wkshps 2011 - Houston, TX, United States
Duration: 05 Dec 201109 Dec 2011

Publication series

Name2011 IEEE GLOBECOM Workshops, GC Wkshps 2011

Conference

Conference2011 IEEE GLOBECOM Workshops, GC Wkshps 2011
Country/TerritoryUnited States
CityHouston, TX
Period05/12/1109/12/11

Fingerprint

Dive into the research topics of 'Downlink beamforming in multi-antenna two-tier networks with user selection'. Together they form a unique fingerprint.

Cite this