Avoiding the Flip Ambiguities in 2D Wireless Sensor Localization by Using Unit Disk Graph Property

TL;DR

This paper introduces a novel 2D wireless sensor localization algorithm that leverages the unit disk graph property to resolve flip ambiguities, especially in cases of limited connectivity, improving localization accuracy.

Contribution

The proposed algorithm uniquely uses the absence of distance measurements alongside existing data to enhance localization precision in wireless sensor networks.

Findings

Improves localization accuracy in low-connectivity scenarios.

Effectively resolves flip ambiguities using unit disk graph properties.

Reduces the number of unlocalized nodes in the network.

Abstract

In this paper, we propose a range-based localization algorithm to localize a wireless sensor network (WSN) in 2D. A widely used algorithm to localize a WSN in 2D is trilateration, which runs in polynomial time. Trilateration uses three distance measurements to localize a node. In some cases, the lack of connectivity leads to a low percentage of localized nodes since a node's position can be fixed using three distance measurements. We propose an algorithm that finds the position of a node by using the absence of a distance measurement in addition to a third distance measurement. If two nodes are not able to sense each other, that means the distance between them is more than the sensing range. Therefore, our algorithm checks if the possible positions of an unlocalized node $u$ is inside the sensing range of a localized node $\ell$ that is not the neighbor of $u$. In such case, we eliminate one of the possible positions.