National Technical University of Athens (NTUA)
Πρόβλεψη και αξιοποίηση ασύρματων διαύλων
Abstract
dc:descriptionThis doctoral thesis deals with the study and development of mechanisms for the predictionand utilization of wireless channels. The enormous increase in the number of spectrumusers, the evolution of wireless services and the necessity for beƩer utilizaƟon has renderedefficient spectral resources management imperaƟve. Cognitive radio networks have beenintroduced to address spectrum scarcity by implementing heterogeneous architectures andadvanced techniques for spectrum access. Cognitive radio networks allow secondary usersto access non-licensed frequency bands or frequency bands already allocated to other users,the so-called primary users.The thesis is divided into three parts. The first part introduces the basic concepts relatedto the thesis and the tools employed to solve the problems related to prediction and utilizationof wireless channels. In the beginning, the key elements of cognitive radio networks andspectrum management are presented as open research issues. Next, the various kinds of spectrum access are presented. A detailed description of various methods for the predictionof primary users traffic performed by secondary users is given in the framework of opportunisticspectrum access. Then, the general concept of spectrum market is discussedfocusing on the different kinds of traded goods, the optimization objectives and the variousforms of spectrum markets. 5G networks and the relevant role of cognitive radio networksis briefly presented. Finally, the basic mathematical tools for tackling problems related tothe predicƟon and utilization of spectrum channels are briefly presented.The second part of the thesis deals with the prediction of wireless channels. The secondaryusers sense the channels unƟl the first channel available is found. Determining the optimalchannel sensing order is crucial for proper exploitation of spectral opportunities. This problemis dealt using reinforcement learning that takes into account information acquired duringthe normal operaƟon of secondary users. This procedure aims at determining the optimalchannel classificaƟon based on (i) their probability of being idle and (ii) their average availabilityduration. The performance of the proposed learning algorithms, which are distributed andbased only on own information, is examined both in staƟc and dynamic environments.The exploitation of wireless channels through spectrum trading is addressed in the thirdpart of the thesis. Enabling hybrid spectrum access of secondary users is investigated. Inthis framework, various system models as well as types of spectrum operators are examined Initially, spectrum trading is approached employing contract theory. Optimal contract designfor traded spectrum and price is determined under complete or incomplete information asto the transmission characteristics of the secondary users. The spectrum trading problem inan environment where different kinds of spectrum operators coexist with secondary usersis addressed with game theory. A spectrum access selection game governing the secondaryusers behaviour and a pricing game played by the spectrum operators are formulated. Nashequilibria are sought for these games and their dependence on the system’s parameters isinvestigated.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2015
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kordali, Angeliki
- Κορδαλή, Αγγελική
Subjects
dc:subject × 27- Γνωστικά δίκτυα επικοινωνιών
- Δυναμική πρόσβαση στο φάσμα
- Ευκαιριακή πρόσβαση στο φάσμα
- Αποκλειστική πρόσβαση στο φάσμα
- Υβριδική πρόσβαση στο φάσμα
- Ανίχνευση φάσματος
- Ενισχυτική μάθηση
- Αγορά φάσματος
- Θεωρία συμβολαίων
- Θεωρία παιγνίων
- Πρόβλημα βελτιστοποίησης
- Cognitive radio networks
- Dynamic spectrum access
- Opportunistic
- Spectrum access
- Exclusive spectrum access
- Hybrid spectrum access
- Spectrum sensing
- contract theory
- Reinforcement learning
- Spectrum trading
- Game theory
- Optimization problem
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/43348