Capacity and coding for quantized MIMO systems

Josef A. Nossek, Michel T. Ivrlač

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

70 Scopus citations

Abstract

In this paper, the effects of receive signal quantization on the channel capacity and the performance of error control coding in multi-input multi-output (MIMO) systems are investigated. The receive antennas of a MIMO system experience a channel-dependent superposition of modulated signals originating from all transmit antennas. A fine-granular analog to digital conversion of the resulting irregular constellation is difficult to obtain in practice. The quantization is therefore likely to be rather coarse. It turns out, however, that the loss in channel capacity due to coarse quantization is surprisingly small. On the other hand, existing coding schemes do not seem to tolerate coarse quantization gracefully, producing rather high error floors. Closing the huge gap between information theoretic opportunities offered by coarse quantized MIMO systems on the one hand, and the actual poor performance of coding schemes which are designed without signal quantization in mind, remains a challenging task.

Original languageEnglish
Title of host publicationIWCMC 2006 - Proceedings of the 2006 International Wireless Communications and Mobile Computing Conference
Pages1387-1391
Number of pages5
DOIs
StatePublished - 2006
EventIWCMC 2006 - 2006 International Wireless Communications and Mobile Computing Conference - Vancouver, BC, Canada
Duration: 3 Jul 20066 Jul 2006

Publication series

NameIWCMC 2006 - Proceedings of the 2006 International Wireless Communications and Mobile Computing Conference
Volume2006

Conference

ConferenceIWCMC 2006 - 2006 International Wireless Communications and Mobile Computing Conference
Country/TerritoryCanada
CityVancouver, BC
Period3/07/066/07/06

Keywords

  • Code-design
  • Information theory
  • MIMO
  • Quantization

Fingerprint

Dive into the research topics of 'Capacity and coding for quantized MIMO systems'. Together they form a unique fingerprint.

Cite this