Performance evaluation of linear constellation precoded OFDM with linear equalizers in overestimated channels

Milad Amir Toutounchian, Sami Muhaidat

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

Abstract

Linear constellation precoded orthogonal frequency division multiplexing (LCP-OFDM) requires a maximum like-lihood (ML) decoder to fully collect multipath diversity gain. However, the complexity of ML increases exponentially with the channel memory, signal constellation size, and the number of transmit/receive antennas. Therefore, linear equalizers are highly desirable. This paper mainly analyzes minimum-mean-square equalization for MMSE-LCP-OFDM systems. We further analyze the performance of linear zero-forcing ZF-LCP-OFDM equalizers. In particular, we show that MMSE-LCP-OFDM equalization provides a significant performance improvement in the so-called overestimated channels. Unlike MMSE-LCP-OFDM, our analysis demonstrates that the performance of ZF-LCP-OFDM is in fact worse in overestimated channels from low to medium SNR region. Simulation results are provided to corroborate the theoretical analysis.

Original languageBritish English
Title of host publication2010 International Conference on Information and Communication Technology Convergence, ICTC 2010
Pages120-124
Number of pages5
DOIs
StatePublished - 2010
Event2010 International Conference on Information and Communication Technology Convergence, ICTC 2010 - Jeju, Korea, Republic of
Duration: 17 Nov 201019 Nov 2010

Publication series

Name2010 International Conference on Information and Communication Technology Convergence, ICTC 2010

Conference

Conference2010 International Conference on Information and Communication Technology Convergence, ICTC 2010
Country/TerritoryKorea, Republic of
CityJeju
Period17/11/1019/11/10

Fingerprint

Dive into the research topics of 'Performance evaluation of linear constellation precoded OFDM with linear equalizers in overestimated channels'. Together they form a unique fingerprint.

Cite this