A hybrid location scheme for sensor networks

Sensor networks consist of so called sensor nodes which comprise sensing, computing, and radio frequency (RF) communication capabilities (see depiction). Used in large numbers, such devices enable sensor networks that can serve in a variety of scenarios-for example, to monitor traffic, metrological changes, catastrophe areas, a particular habitat, or a battlefield. Another possible application is to deploy sensor motes equipped with humidity sensors on an embankment to monitor water heights and warn of flooding.

To accomplish such tasks the nodes have to be context aware. Awareness of the devices' location is one example for that. One can differentiate between numeric (e.g. coordinates) and symbolic (next to the river) location descriptions. Nevertheless there is a interdependence between both: Symbolic regions can basically mapped to coordinates bounding them. A node knowing its coordinates (x,y) and a symbolic region R it resides in could for example spread the information that R comprises at least the point (x,y). If other nodes publish similar information knowledge about the bounds of R would emerge. In such a scenario it can also make sense to use a combination of numeric and symbolic techniques to describe a location.

This master thesis deals with the development of a scheme for describing locations in a both symbolic and numeric way. The scheme should allow for expressing relation such as inside and outside also for mixed location expressions. In addition, the system should enable nodes in the network to learn relations between numeric and symbolic descriptions of the same or related locations.

After studying the relevant literature, approaches should be analytically compared. Promising approaches should be developed into a solution tailored to the constraints of sensor networks which comprise scarce computing power and memory capacity as well as high communication cost. The developed algorithms should be implemented and tested experimentally.

Finally, the state of the art of location schemes, the newly developed algorithm, the implementation and the evaluation results must be documented carefully in the master thesis.