THESIS
2006
xi, 61 leaves : ill. ; 30 cm
Abstract
In wireless sensor networks (WSNs), estimating nodal positions is important for routing efficiency and position-based services. Traditional techniques based on precise measurements of angles and distances are often expensive and power-inefficient. On the other hand, approaches based on landmarks often require many powerful landmarks for information processing....[
Read more ]
In wireless sensor networks (WSNs), estimating nodal positions is important for routing efficiency and position-based services. Traditional techniques based on precise measurements of angles and distances are often expensive and power-inefficient. On the other hand, approaches based on landmarks often require many powerful landmarks for information processing.
In this thesis, we first discuss the applications and traditional techniques to estimate positions in wireless environment. Then a cost-effective and distributed scheme to accurately estimate nodal positions for WSNs is proposed and inves-tigated. It is based on a machine learning tool called ISOMAP and can work with coarse measurements without the need of powerful landmarks. Each node only needs to identify nodes in its neighborhood and exchange information with them to estimate its position accurately. In addition, it allows estimation under network dynamics like nodes joining/leaving the network and presence of nodes with low power. During the position estimation process, useful information for position-based routing is already embedded, so no extra transmission is required for efficient route determination in reporting data to a collecting node.
We have studied and evaluated the scheme’s performance under various sys-tem settings through simulation. It is shown to have fast convergence with low estimation error, even for large networks.
Post a Comment