MU-MIMO channel adapted precoding for MAI/ISI-free uplink burst transmission

This paper considers the uplink of a multiuser direct-sequence code-division multiple-access communication system based on burst transmission over frequency-selective channels. A new precoding method is presented that is able to take benefit from the space diversity introduced at the mobile transmit...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on communications 2004-01, Vol.52 (1), p.100-109
Hauptverfasser: Horlin, F., Vandendorpe, L.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This paper considers the uplink of a multiuser direct-sequence code-division multiple-access communication system based on burst transmission over frequency-selective channels. A new precoding method is presented that is able to take benefit from the space diversity introduced at the mobile transmitters and at the common receiver. User-specific spreading codes are designed so that filters matched to the total impulse responses can remove completely the interference and maximize the signal-and-interference-plus-noise ratios. It is analytically proved that an infinite number of solutions is available to orthogonalize the system. Three particular solutions are considered. A progressive solution is firstly proposed to provide orthogonality in a simple way. Then a near-optimum solution is proposed to orthogonalize the system while attempting to reduce the total transmitted power. The minimization of the total transmitted power is a complex problem of nonlinear optimization that can only be solved using an iterative algorithm. The third solution proposed is based on an iterative algorithm. Those three solutions are analyzed in term of their different performance for a varying number of transmit and receive antennas. A comparison is also provided with a system using a conventional set of binary codes and linear or decision-feedback minimum mean-square error joint detection. It is shown that the near-optimum and iterative solutions are the only ones that are able to fully benefit from the space diversity.
ISSN:0090-6778
1558-0857
DOI:10.1109/TCOMM.2003.822177