JournalspmVol. 68 / No. 1DOI 10.4171/pm/1881

Euclidean distance matrices, semidefinite programming and sensor network localization

  • Abdo Y. Alfakih

    University of Windsor, Windsor, Canada
  • Philippe Charron

    Université de Montréal, Canada
  • Veronica Piccialli

    Università di Roma Tor Vergata, Italy
  • Henry Wolkowicz

    University of Waterloo, Waterloo, Canada
Euclidean distance matrices, semidefinite programming and sensor network localization cover

You need to subscribe to download the article.

Abstract

The fundamental problem of distance geometry involves the characterization and study of sets of points based only on given values of some or all of the distances between pairs of points. This problem has a wide range of applications in various areas of mathematics, physics, chemistry, and engineering. Euclidean distance matrices play an important role in this context by providing elegant and powerful convex relaxations. They play an important role in problems such as graph realization and graph rigidity. Moreover, by relaxing the embedding dimension restriction, these matrices can be used to approximate the hard problems efficiently using semidefinite programming. Throughout this survey we emphasize the interplay between these concepts and problems. In addition, we illustrate this interplay in the context of the sensor network localization problem.