dc.description.abstract | Wireless Sensor Network (WSN) is an important research field in Computer Science
with applications that span multiple domains. Due to the limitation of sensor nodes,
network lifetime is a critical issue that needs to be addressed. Therefore, in this thesis
I propose the Energy-aware Connected k-Neighbourhood (ECKN), a joint position
estimation, packet routing, and sleep scheduling solution that combines some overlap-
ping features. I propose a localization algorithm that performs trilateration using the
position of a mobile sink and of neighbour nodes to estimate the position of a sensor
node with no GPS module. I introduce a routing protocol based on the well-known
Greedy Geographic Forwarding (GGF). Similarly to GGF, my protocol takes into
consideration the position of neighbours to decide the best forwarding node, however
it also considers the residual energy in order to guarantee that the forwarding node
will deliver the packet. The concept of bridges is also introduced, in which the sink
compares its current position with previous positions and calculates whether there is a
shortest path in order to create a bridge that will reduce the number of hops a packet
has to travel through. Lastly, a sleep scheduler is proposed in order to extend the
network lifetime, it is based on the Connected k-Neighbourhood (CKN) algorithm,
which aids in the decision of what nodes goes to sleep while maintaining the network
connected. My sleep scheduler maintains the network denser in the area close to the
sink, since this region receives packets from the whole network to forward to the sink.
An extensive set of performance evaluation experiments is conducted and results show
that ECKN can extend network lifetime, while sustaining acceptable packet delivery
ratio and reducing network overhead. | en |