Aristotle University Of Thessaloniki (AUTH)
Ανάπτυξη μεθοδολογίας με τεχνητά νευρωνικά δίκτυα για την εκτίμηση της απώλειας διαδρομής ραδιοκυμάτων σε δομημένο περιβάλλον ανοιχτού χώρου
Abstract
dc:descriptionThe use of Artificial Neural Networks, so as to predict the propagation path loss in built-up outdoor areas, has been investigated. The novelty of the work at hand emanates from the determination of the kind of information which must be gathered from the built-up environment, according to its local architectural characteristics. That information is later fed to, and utilized from, appropriate Artificial Neural Networks. The effects of parameters such as the operating frequency, the transmitter’s height and its distance from the receiver have been investigated. Two methods of Artificial Neural Network parameter optimization, based on differential evolution, have been proposed in order to make sure that the benefits from the input information are reaped to the biggest extend possible: the former enables the determination of their optimal inner architecture, while the latter deals with optimizing the weights between the synapses, given a predetermined inner structure. Moreover, and in order to compare results with the Artificial Neural Networks, the machine learning method of Random Forests has been implemented in order to obtain path loss predictions. Apart from comparing the results between them, the combination of the two methods in a meta-model in order to obtain even better results, has been proposed.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2018
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Sotiroudis, Sotirios
- Σωτηρούδης, Σωτήριος
Subjects
dc:subject × 16- Διάδοση ραδιοκυμάτων
- Απώλεια διαδρομής
- Τεχνητά νευρωνικά δίκτυα
- Μηχανική μάθηση
- Radio propagation
- Path loss
- Artificial neural networks
- Machine learning
- Φυσικές Επιστήμες
- Φυσική
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Natural Sciences
- Physical Sciences
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
dc:rights- Statement dc:rights
-
- BY_NC_SA
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/45103