Γνωσιακά συστήματα διαχείρισης για δυναμικά δίκτυα πρόσβασης πέραν της 4ης γενιάς

Γεωργακόπουλος, Ανδρέας Ν.

Cognitive management systems for dynamic access networks beyond 4th generation

Doctoral Thesis

Author

Γεωργακόπουλος, Ανδρέας Ν.

Date

2015

Abstract

The vision of a wireless world beyond 4th generation and towards the 5th generation, which will be empowered with cognitive technologies comes closer to reality as traffic traversing through wireless networks gradually increases and novel techniques on cognitive, intelligent management and autonomicity are needed in order to handle effectively the ever-growing traffic demand, diversity and heterogeneity. The opportunistic exploitation of the radio-environment through operator-governed opportunistic networks (ONs) is one of the investigated areas towards the realization of that vision. Operator-governed ONs are dynamically created, temporary, coordinated, device-to-device (D2D) extensions of the infrastructure. Operator governance is being realized through the use of cognitive systems which acquire contextual information of the environment through control channels and make decisions on the creation of ONs. ONs are intended to be operator-governed, in order to assist the infrastructure according to the rules obtained by the operator. Opportunistic use of resources through the establishment of ONs could be used in order to provide extra capacity in congested wireless access network segments e.g., by exploiting the possibility of using neighboring terminals or by utilizing small cells, in order to redirect traffic from congested service areas into non-congested ones. Specifically, this dissertation has investigated the following issues: - Definition of scenarios of ONs and D2D networks for coverage and capacity expansion of the infrastructure through the opportunistic use of neighboring terminals that are available in a service area. These aspects are covered in Chapter 2. - Development of an algorithm for the selection of nodes in order to be part of an ON so as to provide coverage extension of the infrastructure. The algorithm relies on the definition of a fitness function in order to be able to select the ‘good’ neighboring nodes by categorizing them according to certain criteria. These aspects are covered in Chapter 3. - Development of an algorithm for the capacity extension of the infrastructure through the redirection of traffic from congested base stations to neighboring, available base stations. The algorithm is based on Ford–Fulkerson algorithm for computing the maximum flow in a network. These aspects are covered in Chapter 4. - Since the algorithms that have been mentioned previously rely also on the exchange of contextual information between neighboring nodes, it is essential to define certain mechanisms which will ensure the successful exchange of contextual information without adding a lot of overhead to the networks. As a result, Chapter 5 discusses and evaluates the notion of control channels. - Finally, the impact of usage of ONs to energy consumption of the network is investigated by evaluating the benefits of such techniques to the network. These aspects are discussed in Chapter 6. Considering the above structure it can be claimed that the dissertation proceeds to the performance assessment of ONs in situations where coverage and capacity extension of the infrastructure network is deemed necessary. This is actually realized by acquiring contextual information of the neighboring nodes and then selecting the most suitable ones according to certain optimization criteria. Evaluation of various scenarios and use cases is based on extensive simulations in order to increase the validity of the finally extracted conclusions and recommendations. Finally, energy efficiency of such solutions is being assessed in order to check what happens in situations prior to and after the creation of ONs.

Department

Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών Συστημάτων

Number of pages

156

Language

English

URI

https://dione.lib.unipi.gr/xmlui/handle/unipi/8732

Collections

Τμήμα Ψηφιακών Συστημάτων

Show full item record

Except where otherwise noted, this item's license is described as
Attribution-NonCommercial-NoDerivatives 4.0 Διεθνές