RUS  ENG
Full version
JOURNALS // Zhurnal Vychislitel'noi Matematiki i Matematicheskoi Fiziki // Archive

Zh. Vychisl. Mat. Mat. Fiz., 2023 Volume 63, Number 3, Pages 517–530 (Mi zvmmf11531)

Computer science

On stable flows and preflows

A. V. Karzanov

Central Institute of Economics and Mathematics, Russian Academy of Sciences, 117418, Moscow, Russia

Abstract: We propose a new algorithm of finding a stable flow in a network with several sources and sinks. It is based on the idea of preflows (applied in the 1970s for a faster solution of the classical maximal flow problem) and has time complexity $O(nm)$ for a network with $n$ vertices and $m$ edges. The obtained results are further generalized to a larger class of objects, the so-called stable quasi-flows with bounded deviations from the balanced relations in nonterminal vertices.

Key words: stable flow in a network, stable allocation, preflow, quasi-flow.

UDC: 519.86

Received: 11.08.2022
Revised: 26.08.2022
Accepted: 17.11.2022

DOI: 10.31857/S0044466923030079


 English version:
Computational Mathematics and Mathematical Physics, 2023, 63:3, 491–503

Bibliographic databases:


© Steklov Math. Inst. of RAS, 2024