Return to search

High-Rate Space-Time Block Codes in Frequency-Selective Fading Channels

The growing popularity of wireless communications networks has resulted in greater bandwidth contention and therefore spectrally efficient transmission schemes are highly sought after by designers.

Space-time block codes (STBCs) in multiple-input, multiple-output (MIMO) systems are able to increase channel capacity as well as reduce error rate. A general linear space-time structure known as linear dispersion codes (LDCs) can be designed to achieve high-data rates and has been researched extensively for flat fading channels. However, very little research has been done on frequency-selective fading channels. The combination of ISI, signal interference from other transmitters and noise at the receiver mean that maximum likelihood sequence estimation (MLSE) requires high computational complexity. Detection schemes that can mitigate the signal interference can significantly reduce the complexity and allow intersymbol interference (ISI) equalization to be performed by a Viterbi decoder.

In this thesis, detection of LDCs on frequency-selective channels is investigated. Two predominant detection schemes are investigated, namely linear processing and zero forcing (ZF). Linear processing depends on code orthogonality and is only suited for short channels and small modulation schemes. ZF cancels interfering signals when a sufficient number of receive antennas is deployed. However, this number increases with the channel length. Channel decay profiles are investigated for high-rate LDCs to ameliorate this limitation. Performance improves when the equalizer assumes a shorter channel than the actual length provided the truncated taps carry only a small portion of the total channel power.

The LDC is also extended to a multiuser scenario where two independent users cooperate over half-duplex frequency-selective channels to achieve cooperative gain. The cooperative scheme transmits over three successive block intervals. Linear and zero-forcing detection are considered.

Identiferoai:union.ndltd.org:canterbury.ac.nz/oai:ir.canterbury.ac.nz:10092/10360
Date January 2012
CreatorsChu, Alice Pin-Chen
PublisherUniversity of Canterbury. Electrical and Computer Engineering
Source SetsUniversity of Canterbury
LanguageEnglish
Detected LanguageEnglish
TypeElectronic thesis or dissertation, Text
RightsCopyright Alice Pin-Chen Chu, http://library.canterbury.ac.nz/thesis/etheses_copyright.shtml
RelationNZCU

Page generated in 0.0022 seconds