Abstract
This thesis wants to expand the capability of GPS Denied Localisation for human users in areas of multiple elevations, areas where the elevation is more significant compared to its horizontal space, and finding more efficient and accurate node arrangements that could be set up by a human user on the move. We first explore localisation algorithms that retain accuracy even when users travel significant vertical distances compared to horizontal ground. This algorithm works by converting a three-dimension problem into a two-dimension problem, when the user obtains a height reading in the z-axis and the algorithm takes in the height reading for subsequent calculations. We used a barometer to obtain this height reading; the barometer-assisted algorithms were then compared with 3D trilateration for their performance. From this, we examine possible configurations that minimises localisation error while accommodating addition of new anchors. These algorithms rely on geometry between the points to suggest new node locations. The geometry could be applied to implement a localisation network that can expand with new anchors that the user plants while travelling. Both topics of localisation algorithms and node arrangement were examined using simulations and real time localisation field trials. The noise models within the simulations were derived from experimental tests on the sensors used within the field trials. We wish to apply the techniques and results of this research for human users who need to enter unexplored areas with no prior knowledge, do not have GPS support, and cannot assume prior support from others.