A Low Complexity Massive MIMO Detection Scheme Using Angular-Domain Processing

Forskningsoutput: Kapitel i bok/rapport/Conference proceedingKonferenspaper i proceeding

Abstract

Signal processing complexity and required memory become problematic in massive MIMO systems as the dimension of channel state information (CSI) matrix grows significantly with the large number of antennas and users. To address these challenges, we propose the first angular-domain massive MIMO detection scheme, which is based on three concepts: transferring the baseband processing from the spatial domain to the angular domain; exploiting the sparsity of received beams to reduce the dimension of CSI matrix; and performing the whole detection and precoding in the angular domain using the reduced CSI matrix. We have measured the massive MIMO channel at 2.6 GHz with a 128-antenna linear array communicating with 16 users to evaluate our scheme. Complexity analysis and simulations show that proposed idea leads to 40% – 70% reduction in the processing complexity and memory without significant performance loss, which significantly outperforms the antenna-domain schemes.

Detaljer

Författare
Enheter & grupper
Forskningsområden

Ämnesklassifikation (UKÄ) – OBLIGATORISK

  • Annan elektroteknik och elektronik
  • Kommunikationssystem

Nyckelord

Originalspråkengelska
Titel på värdpublikationIEEE Global Conference on Signal and Information Processing (GlobalSIP)
FörlagIEEE - Institute of Electrical and Electronics Engineers Inc.
Sidor181-185
Antal sidor5
ISBN (elektroniskt)978-1-7281-1295-4
ISBN (tryckt)978-1-7281-1296-1
StatusPublished - 2018
PublikationskategoriForskning
Peer review utfördJa
Evenemang2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP) - Anaheim, Anaheim, USA
Varaktighet: 2018 nov 262018 nov 29

Konferens

Konferens2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)
LandUSA
OrtAnaheim
Period2018/11/262018/11/29