Skip to main navigation Skip to search Skip to main content

Efficient CSI Correction with CFO for Uplink MU-MIMO Networks

  • Xi'an Jiaotong University

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

3 Scopus citations

Abstract

Multiuser MIMO (MU-MIMO) plays an important role in Wireless Local Area Networks (WLAN), due to its ability of enabling concurrent transmissions between multiple users and Access Points. However, a larger amount of erroneous packets often occur in the MU-MIMO streams, resulting in massive retransmissions and severe transmission lags. Based on our empirical observations, we found that the incorrect feedback of Channel State Information (CSI, which is used for encoding and decoding packets in MU-MIMO) is a main reason to yield erroneous packets. Unfortunately, prior researches mainly focus on the data errors caused by low SNR, while paying little attention to those caused by incorrect CSI. In this paper, we ascertain the carrier frequency offset (CFO) causing the CSI errors, and show its impacts on the MU-MIMO uplink transmission. To solve the problem, we propose an efficient CSI correction with CFO, namely ECC. ECC quickly identifies and corrects CSI errors, and hence significantly reinforces the efficiency of MU-MIMO. ECC is practical and fully compatible with current 802.11ac WLAN frameworks. We implement ECC over USRP platforms. Extensive experiment results demonstrate that ECC substantially improves the throughput for MU-MIMO WLAN systems, (e.g., more than 5X for 2X2 uplink MU-MIMO).

Original languageEnglish
Title of host publicationProceedings - 2017 3rd International Conference on Big Data Computing and Communications, BigCom 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages256-264
Number of pages9
ISBN (Electronic)9781538633496
DOIs
StatePublished - 15 Nov 2017
Event3rd International Conference on Big Data Computing and Communications, BigCom 2017 - Chengdu, Sichuan, China
Duration: 10 Aug 201711 Aug 2017

Publication series

NameProceedings - 2017 3rd International Conference on Big Data Computing and Communications, BigCom 2017

Conference

Conference3rd International Conference on Big Data Computing and Communications, BigCom 2017
Country/TerritoryChina
CityChengdu, Sichuan
Period10/08/1711/08/17

Fingerprint

Dive into the research topics of 'Efficient CSI Correction with CFO for Uplink MU-MIMO Networks'. Together they form a unique fingerprint.

Cite this