Temporal flows in temporal networks

Eleni C. Akrida, Jurek Czyzowicz, Leszek Gasieniec, Lukasz Kuszner, Paul G. Spirakis

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Scopus citations


We introduce temporal flows on temporal networks [17, 19], i.e., networks the links of which exist only at certain moments of time. Such networks are ephemeral in the sense that no link exists after some time. Our flow model is new and differs from the “flows over time” model, also called “dynamic flows” in the literature. We show that the problem of finding the maximum amount of flow that can pass from a source vertex s to a sink vertex t up to a given time is solvable in Polynomial time, even when node buffers are bounded. We then examine mainly the case of unbounded node buffers. We provide a simplified static Time-Extended network (STEG), which is of polynomial size to the input and whose static flow rates are equivalent to the respective temporal flow of the temporal network; using STEG, we prove that the maximum tem-poral flow is equal to the minimum temporal s-t cut. We further show that temporal flows can always be decomposed into flows, each of which moves only through a journey, i.e., a directed path whose successive edges have strictly increasing moments of existence. We partially char-acterise networks with random edge availabilities that tend to eliminate the s → t temporal flow. We then consider mixed temporal networks, which have some edges with specified availabilities and some edges with random availabilities; we show that it is #P-hard to compute the tails and expectations of the maximum temporal flow (which is now a random variable) in a mixed temporal network.

Original languageEnglish (US)
Title of host publicationAlgorithms and Complexity - 10th International Conference, CIAC 2017, Proceedings
EditorsDimitris Fotakis, Aris Pagourtzis, Vangelis Th. Paschos
PublisherSpringer Verlag
Number of pages12
ISBN (Print)9783319575858
StatePublished - 2017
Externally publishedYes
Event10th International Conference on Algorithms and Complexity, CIAC 2017 - Athens, Greece
Duration: May 24 2017May 26 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10236 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference10th International Conference on Algorithms and Complexity, CIAC 2017

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)


Dive into the research topics of 'Temporal flows in temporal networks'. Together they form a unique fingerprint.

Cite this