National Technical University of Athens (NTUA)
Βελτιστοποίηση ασύρματων κυψελωτών δικτύων με υποστήριξη άμεσης επικοινωνίας
Abstract
dc:descriptionThis PhD thesis addresses the problem of optimal resource allocation in Device to Device (D2D) enabled cellular networks. In this type of networks users may communicate directly with each other either by using different resource blocks (RBs) or sharing common RBs with the conventional cellular users. The general perspective of the thesis is to allocate RBs and transmission power to users in order to satisfy their minimum signal to interference and noise ratio (SINR) constraints with the minimum transmitted power.For the overlay case, a queueing system is proposed in order to model the RB allocation process. Our aim is to balance and minimize the blocking probability of each type of user of the network. This model finds the optimal policy for the RB allocation and a heuristic algorithm is used in order to find the necessary parameters of the optimal policy, which is shown to be of threshold type.Then, we focus on the underlay case of RB sharing where cellular and D2D users share the same RBs. A game theoretic approach is proposed for the power control problem for a given RB. A Nash equilibrium (NE) point is proven to exist and a best response algorithm is proposed in order to converge to the NE point of the game. In this way, the users of the network find an optimal tradeoff between the amount of transmitted power and the required SINR level at their receivers.Next, we formulate the problem of interference minimization under users’ SINR constraints for a D2D underlay cellular network. We study both the single cell and the multicell case. Due to the computational intractability of the problem, it is decomposed by two separate subproblems. The first one is a RB allocation problem under fixed transmission power while the second, following the result from the first one, is a power control problem for every shared RB. For both problems, efficient algorithms are proposed. The first problem is modelled as an exact potential game where the total interference of the network coincides the opposite of the potential function of the respective game. Leveraging the desirable properties of the exact potential games a best response algorithm is proposed which converges to the NE point of the respective game. After that, the power control problem is modelled as a concave game where a distributed algorithm for the NE computation is proposed. This algorithm is proven to converge to the NE of the game after finite number of iterations.The previous approach is further extended in order to cover the case of multicell D2D underlay cellular network. In the first step, the total interference minimization problem under fixed transmission power is modelled as a bilateral symmetric game which belongs to a special category of exact potential games. Thus, following a best response algorithm, we reach a NE point of the game under finite number of iterations. Moreover, a random distributed version of the best response algorithm is proposed which achieves fair performance with significantly lower number of iterations and communication overhead. Then the problem of interference minimization for the RB allocation vector proposed by the first stage is shown to be equivalent to a total power minimization problem under the same constraints. Following the well-known Foschini Miljanic algorithm which solves the total power minimization problem a suitable solution for our problem is obtained. Finally, we demonstrate via extensive numerical simulations that the total transmitted power is reduced dramatically and the number of users who satisfy their SINR requirements increases significantly.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2017
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Katsinis, Georgios
- Κατσίνης, Γεώργιος
Subjects
dc:subject × 14- Ασύρματα δίκτυα
- Διαχείριση παρεμβολών
- Θεωρία παιγνίων
- Ανάθεση πόρων
- Άμεση επικοινωνία των χρηστών
- D2D enabled wireless cellular networks
- Interference management
- Resource allocation
- Game theory
- Potential games
- 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/42012