National and Kapodistrian University of Athens
Μη γραμμική επεξεργασία σήματος με εφαρμογές στις τηλεπικοινωνίες
Abstract
dc:descriptionNonlinear behaviour appears in almost all digital communication systems. Linear approximations that do not take into account this type of behaviour, may lead to system performance degradation as well as loss of information. Therefore, appropriate models should be developed that tackle nonlinear system characteristics. The primary objective of this dissertation is the solution of the structure determination problem in system identification. We develop methods and algorithms for the order determination of both linear and nonlinear systems and also for system identification and equalization. The methodology developed in this thesis builds upon subspaces induced by projecting system output data into past and future values. The concept is based in smoothing techniques. Specifically, a window of specific length is chosen and data values that fall within window are orthogonally projected into past and future values. Then, the error projection matrix is developed and rank detection techniques are employed to identify subsystem orders. Having identified discrete subsystem orders, we solve the projection error matrix equation for each group of subsystems and either identify a linear channel or equalize non-linear channels.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2012
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kotoulas, Diamantis
- Κωτούλας, Διαμαντής
Subjects
dc:subject × 14Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/35355