Robust Positioning Algorithms for Distributed Ad-Hoc Wireless Sensor Networks

TLDR

The paper presents a distributed algorithm for determining node positions in ad‑hoc wireless sensor networks. The algorithm is implemented and evaluated on networks of 400 randomly placed nodes, with error magnitudes expressed as percentages of each node’s radio range. The algorithm achieves average errors below 33 % under 5 % distance errors with 5 % anchors and 7‑node connectivity, and remains effective with up to 40 % distance errors when connectivity is at least 12 nodes and 10 % anchors.

Abstract

A distributed algorithm for determining the positions of nodes in an ad-hoc, wireless sensor network is explained in detail. Details regarding the implementation of such an algorithm are also discussed. Experimentation is performed on networks containing 400 nodes randomly placed within a square area, and resulting error magnitudes are represented as percentages of each node’s radio range. In scenarios with 5% errors in distance measurements, 5% anchor node population (nodes with known locations), and average connectivity levels between neighbors of 7 nodes, the algorithm is shown to have errors less than 33% on average. It is also shown that, given an average connectivity of at least 12 nodes and 10% anchors, the algorithm performs well with up to 40% errors in distance measurements.