Sparse fault-tolerant BFS trees

Merav Parter, David Peleg

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

39 Scopus citations

Abstract

A fault-tolerant structure for a network is required to continue functioning following the failure of some of the network's edges or vertices. This paper considers breadth-first search (BFS) spanning trees, and addresses the problem of designing a sparse fault-tolerant BFS tree, or FT-BFS tree for short, namely, a sparse subgraph T of the given network G such that subsequent to the failure of a single edge or vertex, the surviving part T′ of T still contains a BFS spanning tree for (the surviving part of) G. For a source node s, a target node t and an edge e ∈ G, the shortest s - t path P s,t,e that does not go through e is known as a replacement path. Thus, our FT-BFS tree contains the collection of all replacement paths P s,t,e for every t ∈ V(G) and every failed edge e ∈ E(G). Our main results are as follows. We present an algorithm that for every n-vertex graph G and source node s constructs a (single edge failure) FT-BFS tree rooted at s with O(n · min{Depth(s), √n}) edges, where Depth(s) is the depth of the BFS tree rooted at s. This result is complemented by a matching lower bound, showing that there exist n-vertex graphs with a source node s for which any edge (or vertex) FT-BFS tree rooted at s has Ω(n3/2) edges. We then consider fault-tolerant multi-source BFS trees, or FT-MBFS trees for short, aiming to provide (following a failure) a BFS tree rooted at each source s ∈ S for some subset of sources S ⊆ V. Again, tight bounds are provided, showing that there exists a poly-time algorithm that for every n-vertex graph and source set S ⊆ V of size σ constructs a (single failure) FT-MBFS tree T*(S) from each source si ∈ S, with O(√σ·n3/2) edges, and on the other hand there exist n-vertex graphs with source sets S ⊆ V of cardinality σ, on which any FT-MBFS tree from S has Ω(√σ·n3/2) edges. Finally, we propose an O(logn) approximation algorithm for constructing FT-BFS and FT-MBFS structures. The latter is complemented by a hardness result stating that there exists no Ω(logn) approximation algorithm for these problems under standard complexity assumptions. In comparison with previous constructions our algorithm is deterministic and may improve the number of edges by a factor of up to √n for some instances. All our algorithms can be extended to deal with one vertex failure as well, with the same performance.

Original languageEnglish
Title of host publicationAlgorithms, ESA 2013 - 21st Annual European Symposium, Proceedings
Pages779-790
Number of pages12
DOIs
StatePublished - 2013
Externally publishedYes
Event21st Annual European Symposium on Algorithms, ESA 2013 - Sophia Antipolis, France
Duration: 2 Sep 20134 Sep 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8125 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference21st Annual European Symposium on Algorithms, ESA 2013
Country/TerritoryFrance
CitySophia Antipolis
Period2/09/134/09/13

Bibliographical note

Funding Information:
Supported in part by the Israel Science Foundation (grant 894/09), the I-CORE program of the Israel PBC and ISF (grant 4/11), the United States-Israel Binational Science Foundation (grant 2008348), the Israel Ministry of Science and Technology (infrastructures grant), and the Citi Foundation.

Fingerprint

Dive into the research topics of 'Sparse fault-tolerant BFS trees'. Together they form a unique fingerprint.

Cite this