Design of unitary precoders for ISI channels

Redundancy introduced using filterbank precoders at the transmitter builds a unified framework for modulation schems. Taking advantage of this diversity can offer a powerful tool for removing interblock interference and devising simple but effective precoders for suppressing the intersymbol interfer...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2002 IEEE International Conference on Acoustics, Speech, and Signal Processing Jg. 3; S. III-2265 - III-2268
Hauptverfasser: Zhang, Jian-Kang, Kavcic, Aleksandar, Ma, Xiao, Wong, Kon Max
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.01.2002
Schlagworte:
Artificial intelligence
Baseband
Equalizers
Matrix decomposition
Multiplexing
Recurrent neural networks
Silicon carbide
ISBN:9780780374027, 0780374029
ISSN:1520-6149
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Redundancy introduced using filterbank precoders at the transmitter builds a unified framework for modulation schems. Taking advantage of this diversity can offer a powerful tool for removing interblock interference and devising simple but effective precoders for suppressing the intersymbol interference (ISI) and being robust to frequency selective channels.. In this paper, we assume that the transmitter knows the autocorrelation sequences of the channel impulse response. Under this assumption, we derive a lower and an upper bound on the free distance for a precoded channel and with this, design the precoder that maximizes the lower bound subject to a power constraint. It turns out that the optimal precoder is a unitary matrix which makes QR decomposition of its super-channel exhibit equal diagonal entries in R-factor and the lower and the upper bounds of its free distance equal. We show that for the optimal precoded channel, the detection performance using the decision feedback equalizer (DFE) based on QR decomposition is asymptotically equivalent to that of the mamximum likehood detector (MLD) when the signal to noise ratio (SNR) is large.
ISBN:9780780374027
0780374029
ISSN:1520-6149
DOI:10.1109/ICASSP.2002.5745096