Schlüsselwörter
(Englisch)
|
Wireless sensors; ad hoc networks; dynamical models; adaptive energy aware algorithms
|
Forschungsprogramme
(Englisch)
|
COST-Action 295 - Dynamic Communication Networks: Foundations and Algorithms
|
Kurzbeschreibung
(Englisch)
|
The focus of this project is on devising models for dynamical aspects of wireless sensor networks. Indeed, the specific characteristics of the media of transmission for wireless communications, for instance radio communications, raise new issues to be handled. Compared to classical communications networks interferences are much more malicious and difficult to handle since they are usually undetectable. Hence, statistical knowledge of the state of the communication channels must be predicted, based on dynamical models, to design reliable communication networks and protocols. Apart from the statistical use of the channels an important parameter is the relative positions of the stations. Indeed, for fixed power of emission, the topological properties of the communication graph rely completely on these relative positions as well as the noise generated by simultaneous transmission. Moreover, results on percolation in the communication graphs show that the geometry of the region in which are scattered the sensors plays a key role as well as the radiation emission pattern. These spatial data have to be included as characteristics of the sensor networks to ensure a good quality of the models predictions. Such models are relevant to evaluate the quality of services (QoS), efficiency and reliability of wireless networks. Models are based on statistical hypothesis on the channel occupations by the stations composing the networks, are probabilistic in nature and usually model a typical application of the networks. Issues related to the dynamic ad hoc deployment of the networks as well as management of faulty stations or mobility of sensors are to be considered. Moreover, typical applications of sensor networks need to adapt to the external conditions in order to provide more efficient service or to take care of classical issue such as energy consumption, for instance by implementing adaptive routing algorithms.
|
Weitere Hinweise und Angaben
(Englisch)
|
Full name of research-institution/enterprise: Université de Genève Centre Universitaire d'Informatique C.U.I. Battelle, Bât. A
|
Partner und Internationale Organisationen
(Englisch)
|
BE, CH, CY, CZ, DE, EE, ES, FI, FR, GR, HU, IL, IT, NL, NO, PL, PT, SE, SI, UK
|
Abstract
(Englisch)
|
In this research project we focus on dynamical model of processes evolving on the top of sensor networks. We consider gathering data by a single sink where data are generated by sensors and conveyed either in a hop-by-hop way or directly sent to the sink. The difference of energy consumption between the different path is used to balance the total energy consumption among the entire set of sensors. The energy-balance mechanism is based on simple rules which need only local knowledge of the network topology and sensors activity. We establish results concerning the existence of such mechanisms, the equivalence between balancing the consumed energy and making the lifetime of the network the longer and, optimal results relative to the flow of data in the network. We also consider the selection of optimal paths of data propagation using local mechanism with no traffic overhead. The mechanisms we suggest and analyse are based on the use of trust mechanism. More precisely, we consider greedy forwarding which consist in forwarding data to the node which is the closer to the destination among the set of neighbours. If greedy fails because there is no such sensor, perimeter routing is used. Because perimeter routing is more costly in term of energy and not optimal in terms of hop distance, we define the nodes which use perimeter routing as untrustworthy. As all the nodes are classified as trusty or untrustworthy, greedy routing consider only trusty nodes and the resulting paths shown to be close to the optimal ones.
|
Datenbankreferenzen
(Englisch)
|
Swiss Database: COST-DB of the State Secretariat for Education and Research Hallwylstrasse 4 CH-3003 Berne, Switzerland Tel. +41 31 322 74 82 Swiss Project-Number: C05.0030
|
