BD-CAT: Balanced dynamic content addressing in trees

Research output: Contribution to book/Conference proceedings/Anthology/ReportConference contributionContributedpeer-review

Contributors

Abstract

Balancing the load in content addressing schemes for route-restricted networks represents a challenge with a wide range of applications. Solutions based on greedy embeddings maintain minimal state information and enable efficient routing, but any such solutions currently result in either imbalanced content addressing, overloading individual nodes, or are unable to efficiently account for network dynamics. In this work, we propose a greedy embedding in combination with a content addressing scheme that provides balanced content addressing while at the same time enabling efficient stabilization in the presence of network dynamics. We point out the tradeoff between stabilization complexity and maximal permitted imbalance when deriving upper bounds on both metrics for two variants of the proposed algorithms. Furthermore, we substantiate these bounds through a simulation study based on both real-world and synthetic data.

Details

Original languageEnglish
Title of host publicationIEEE INFOCOM 2017 - IEEE Conference on Computer Communications
PublisherWiley-IEEE Press
Pages1-9
Number of pages9
ISBN (print)978-1-5090-5337-7
Publication statusPublished - 4 May 2017
Peer-reviewedYes

Conference

Title2017 IEEE Conference on Computer Communications
Abbreviated titleIEEE INFOCOM 2017
Duration1 - 4 May 2017
CityAtlanta
CountryUnited States of America

External IDs

Scopus 85034026278

Keywords

Keywords

  • Routing, Topology, Heuristic algorithms, Content addressable storage, Complexity theory, Conferences, Network topology