Συντάχθηκε 17-09-2010 14:04
από Eleni Stamataki
Email συντάκτη: estamataki<στο>tuc.gr
Ενημερώθηκε:
-
Ιδιότητα: σύνταξη/αποχώρηση υπάλληλος.
ΤΜΗΜΑ ΗΛΕΚΤΡΟΝΙΚΩΝ ΜΗΧΑΝΙΚΩΝ ΚΑΙ ΜΗΧΑΝΙΚΩΝ ΥΠΟΛΟΓΙΣΤΩΝ
ΠΑΡΟΥΣΙΑΣΗ ΔΙΔΑΚΤΟΡΙΚΗΣ ΔΙΑΤΡΙΒΗΣ
“Performance degradation of telecommunication transceivers due to parameter estimation errors”
ΤΣΙΠΟΥΡΙΔΟΥ ΔΕΣΠΟΙΝΑ
Τετάρτη 22 Σεπτεμβρίου 2010, Ώρα 13:00
Αμφιθέατρο Κτ. Επιστημών, Πολυτεχνειούπολη
Εξεταστική Επιτροπή: Καθ. Αθανάσιος Λιάβας, Τμήμα ΗΜΜΥ, Πολυτεχνείου Κρήτης
Καθ. Νικόλαος Σιδηρόπουλος, Τμήμα ΗΜΜΥ, Πολυτεχνείου Κρήτης
Καθ. Βασίλειος Διγαλάκης, Τμήμα ΗΜΜΥ, Πολυτεχνείου Κρήτης
Καθ. Μιχαήλ Πατεράκης, Τμήμα ΗΜΜΥ, Πολυτεχνείου Κρήτης
Επ. Καθ. Γεώργιος Καρυστινός, Τμήμα ΗΜΜΥ, Πολυτεχνείου Κρήτης
Καθ. Σέργιος Θεοδωρίδης, Τμήμα Πληροφορικής & Τηλεπικοινωνιών, ΕΚΠΑ
Καθ.Κων/νος Μπερμπερίδης, Τμήμα Μηχ. Η/Υ και Πληροφορικής, Παν. Πατρών
ΠΕΡΙΛΗΨΗ
In this thesis we consider the performance degradation of three telecommunication transceivers with respect to parameter uncertainties. The major contribution of our analysis is the derivation of simple and informative approximations to the excess mean square error, which uncover the basic factors that determine the performance of each transceiver under the presence of parameter uncertainties.
In the first part of the work, we consider the case of the Multiple-Input Multiple-Output (MIMO) flat fading channel and two widely known transceiver schemes, i.e., the transmit MIMO Wiener filtering and the Tomlinson-Harashima (TH) precoding. For the case of the transmit MIMO Wiener filtering, degradation is due to channel estimation errors and time-variations, and noise second-order statistics estimation errors. For the case of the TH precoder, degradation is due to channel estimation errors and time-variations. For both cases, our final expressions uncover the factors that determine the performance degradation in practice, and the relative importance between different error sources.
In the second part of the work, we consider the case of the Single-Input Single-Output (SISO) frequency selective channel with the minimum mean square error (MMSE) linear equalizer. The error sources that cause the system degradation is the channel and CFO estimation errors. Our aim is to uncover the relative importance of these error sources. It turns out that the CFO estimation error is much more important than the channel estimation error. This fact leads to useful conclusions concerning the optimal training sequence design for joint CFO and channel estimation.
