Effects on the algebraic connectivity of weighted graphs under edge rotations

Xinzhuang Chen, Shenggui Zhang, Shanshan Gao, Xiaodi Song

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

For a weighted graph G, the rotation of an edge uv1 from v1 to a vertex v2 is defined as follows: delete the edge uv1, set w(uv2) as w(uv1)+w(uv2) if uv2 is an edge of G; otherwise, add a new edge uv2 and set w(uv2)=w(uv1), where w(uv1) and w(uv2) are the weights of the edges uv1 and uv2, respectively. In this paper, effects on the algebraic connectivity of weighted graphs under edge rotations are studied. For a weighted graph, a sufficient condition for an edge rotation to reduce its algebraic connectivity and a necessary condition for an edge rotation to improve its algebraic connectivity are proposed based on Fiedler vectors of the graph. As applications, we show that, by using a series of edge rotations, a pair of pendent paths (a pendent tree) of a weighted graph can be transformed into one pendent path (pendent edges attached at a common vertex) of the graph with the algebraic connectivity decreasing (increasing) monotonically. These results extend previous findings of reducing the algebraic connectivity of unweighted graphs by using edge rotations.

Original languageEnglish
Pages (from-to)289-301
Number of pages13
JournalLinear Algebra and Its Applications
Volume703
DOIs
StatePublished - 15 Dec 2024

Keywords

  • Algebraic connectivity
  • Edge rotation
  • Fiedler vector
  • Pendent tree (path)

Fingerprint

Dive into the research topics of 'Effects on the algebraic connectivity of weighted graphs under edge rotations'. Together they form a unique fingerprint.

Cite this