Aristotle University Of Thessaloniki (AUTH)
Σχεδιασμός ασύρματων δικτύων με δυνατότητα παροχής προηγμένων υπηρεσιών
Abstract
dc:descriptionIn this dissertation, three optimization problems in the context of wireless ad-hoc networks are solved. Specifically, a dual-hop system with non-regenerative relay under arbitrary fading is examined, where the relay is subject to instantaneous as well as average output power constraints. The problem posed is how to select the relay gain, as a function of channel state, in order to maximize a suitably defined receiver utility function, which contains, as special cases, the Shannon rate, outage probability and complementary symbol error probability of typical modulation schemes. The analysis, using tools from convex analysis and combinatorial optimization, is mostly theoretical and is easily extended to dual-branch diversity systems, of either selection (SC) or maximal ratio combining (MRC) type. The next two problems are related to battery-operated sensor networks. In the first scenario, the node batteries are non-rechargeable and the nodes need to send the generated information to a mobile sink, which is allowed to move among a set of specified locations. All links are assumed to be deterministic and a fluid model is used for the packet traffic. The question is what is the sink’s sojourn time at each allowable location and what is the corresponding routed flow so that the network “lifetime” is maximized. The latter term, which has numerous definitions in the literature, is defined in this dissertation as the time up to the (earliest) battery depletion of any node. Imposing energy and power constraints to all nodes (except for the sink), a linear program is formulated. Using analytical techniques, its dual problem is solved via an iterative subgradient projection procedure, at each step of which minimum cost flow problems must be solved, and an algorithm that is amenable to distributed implementation (a major objective in ad-hoc networks) is proposed. Finally, the scenario of rechargeable batteries is also examined, where the network lifetime is sufficiently large (in the order of many years) so that it can be approximately regarded as infinite. This shifts the design focus towards optimal network performance. The latter is quantified through a utility function of the long-term rates of the bits admitted into the network. Also, due to the infinite lifetime, stability of network queues is a crucial issue that must be addressed. Using stochastic models for the arrival, channel state and battery recharge processes, an online policy that stabilizes the network, with a guaranteed achieved performance lower bound, is proposed. The policy requires knowledge of current values only and is asymptotically optimal as, where is the maximal battery capacity and is the maximal allowable transmission power per slot. Simulations are performed for the above three problems and the obtained results demonstrate the feasibility of the proposed methods and algorithms.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Gatzianas, Marios-Apostolos
- Γκατζιάνας, Μάριος-Απόστολος
Subjects
dc:subject × 16- Ασύρματα δίκτυα
- Δίκτυα αισθητήρων
- Έλεγχος δικτύων
- Μη-αναγεννητικοί αναμεταδότες
- Χρόνος ζωής δικτύου
- Επαναφορτιζόμενες μπαταρίες
- Wireless networks
- Sensor networks
- Network control
- Non-regenerative relays
- Network lifetime
- Rechargeable batteries
- 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/20718