Analysis of Time and Frequency Domain Pace Algorithms for OFDM with Virtual Subcarriers

Christian Rom, Carles Navarro Manchón, Luc Deneire, Troels Bundgaard Sørensen, Preben Mogensen

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

2 Citations (Scopus)

Abstract

This paper studies common linear frequency direction pilot-symbol aided channel estimation algorithms for orthogonal frequency division multiplexing in a UTRA long term evolution context. Three deterministic algorithms are analyzed: the maximum likelihood (ML) approach, the noise reduction algorithm (NRA) and the robust Wiener (RW) filter. A closed form mean squared error is provided for these three algorithms. Analytical and simulation results show that, in the presence of virtual subcarriers, the ML can suffer large performance degradation due to ill-conditioned matrix issues. A solution to this problem is to use the Tikhonov regularization method giving the NRA. The equivalence between the NRA and the RW is proved analytically. A practical implementation of the NRA and RW is proposed based on partial-input partial-output FFT, leading to 6 to 8 times lower complexity than the reference implementation.
Original languageEnglish
Title of host publicationIkke angivet
Number of pages5
PublisherElectrical Engineering/Electronics, Computer, Communications and Information Technology Association
Publication date2007
ISBN (Electronic)978-1-4244-1144-3
DOIs
Publication statusPublished - 2007
EventIEEE International Symposium on Personal, Indoor and Mobile Radio Communications. PIMRC 2007 - Athens, Greece
Duration: 3 Sept 20077 Sept 2007
Conference number: 18

Conference

ConferenceIEEE International Symposium on Personal, Indoor and Mobile Radio Communications. PIMRC 2007
Number18
Country/TerritoryGreece
CityAthens
Period03/09/200707/09/2007

Fingerprint

Dive into the research topics of 'Analysis of Time and Frequency Domain Pace Algorithms for OFDM with Virtual Subcarriers'. Together they form a unique fingerprint.

Cite this