On the Inefficiency of Atomic Splittable Routing Games over Parallel Links

Abstract Several recent works on non-atomic routing games suggest that the performance degradation of selfish routing with respect to optimal routing is overall low and far from worst-case scenarios. In this work, we study the performance degradation induced by the lack of coordination in an atomic routing game over parallel links in which there are two types of links. The latency function of "cheap" links is of the form $$c_1 \phi (x)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>1</mml:mn> </mml:msub> <mml:mi>ϕ</mml:mi> <mml:mrow> <mml:mo>(</mml:mo> <mml:mi>x</mml:mi> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:math> , whereas the latency function of "expensive" links is of the form $$c_2 \phi (x)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>2</mml:mn> </mml:msub> <mml:mi>ϕ</mml:mi> <mml:mrow> <mml:mo>(</mml:mo> <mml:mi>x</mml:mi> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:math> , where $$c_2&gt;c_1$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>2</mml:mn> </mml:msub> <mml:mo>&gt;</mml:mo> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>1</mml:mn> </mml:msub> </mml:mrow> </mml:math> . We obtain an explicit characterization of the optimal and equilibrium flow configurations, and establish sufficient conditions on the latency function $$\phi (x)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>ϕ</mml:mi> <mml:mo>(</mml:mo> <mml:mi>x</mml:mi> <mml:mo>)</mml:mo> </mml:mrow> </mml:math> under which the worst traffic conditions occur when all users have the same traffic demand and the total traffic demand is such that "expensive" link are marginally used by selfish routing. We also obtain some partial results on the worst network configuration for the inefficiency of selfish routing. All in all, our results suggest that the worst-case scenario for the inefficiency of selfish routing corresponds to very specific traffic conditions and to highly asymmetric network configurations, and thus that the Price of Anarchy is probably an overly pessimistic performance measure for non-cooperative routing games, as advocated in the above-mentioned works.

Authors

Institutions

Publication Details

Journal
Annals of Operations Research
Published
2025-05-24
DOI
https://doi.org/10.1007/s10479-025-06648-y
Citations
1
Primary Topic
Quantum Computing Algorithms and Architecture
Type
article
Field-Weighted Citation Impact
2.48

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

On the Inefficiency of Atomic Splittable Routing Games over Parallel Links

Olivier Brun, Josu Doncel
1 citations
Annals of Operations Research
Quantum Computing Algorithms and Architecture
2.48
article

On the Inefficiency of Atomic Splittable Routing Games over Parallel Links

Olivier Brun, Josu Doncel
article en
1 citations

Abstract

Abstract Several recent works on non-atomic routing games suggest that the performance degradation of selfish routing with respect to optimal routing is overall low and far from worst-case scenarios. In this work, we study the performance degradation induced by the lack of coordination in an atomic routing game over parallel links in which there are two types of links. The latency function of "cheap" links is of the form $$c_1 \phi (x)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>1</mml:mn> </mml:msub> <mml:mi>ϕ</mml:mi> <mml:mrow> <mml:mo>(</mml:mo> <mml:mi>x</mml:mi> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:math> , whereas the latency function of "expensive" links is of the form $$c_2 \phi (x)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>2</mml:mn> </mml:msub> <mml:mi>ϕ</mml:mi> <mml:mrow> <mml:mo>(</mml:mo> <mml:mi>x</mml:mi> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:math> , where $$c_2&gt;c_1$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>2</mml:mn> </mml:msub> <mml:mo>&gt;</mml:mo> <mml:msub> <mml:mi>c</mml:mi> <mml:mn>1</mml:mn> </mml:msub> </mml:mrow> </mml:math> . We obtain an explicit characterization of the optimal and equilibrium flow configurations, and establish sufficient conditions on the latency function $$\phi (x)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>ϕ</mml:mi> <mml:mo>(</mml:mo> <mml:mi>x</mml:mi> <mml:mo>)</mml:mo> </mml:mrow> </mml:math> under which the worst traffic conditions occur when all users have the same traffic demand and the total traffic demand is such that "expensive" link are marginally used by selfish routing. We also obtain some partial results on the worst network configuration for the inefficiency of selfish routing. All in all, our results suggest that the worst-case scenario for the inefficiency of selfish routing corresponds to very specific traffic conditions and to highly asymmetric network configurations, and thus that the Price of Anarchy is probably an overly pessimistic performance measure for non-cooperative routing games, as advocated in the above-mentioned works.

Annals of Operations ResearchVol. 356(1)
Centre National de la Recherche Scientifique (FR), University of the Basque Country (ES), Université Fédérale de Toulouse Midi-Pyrénées (FR), Laboratoire d'Analyse et d'Architecture des Systèmes (FR), Institut National des Sciences Appliquées de Toulouse (FR), Toulouse Mathematics Institute (FR)
Hezkuntza, Hizkuntza Politika Eta Kultura Saila, Eusko Jaurlaritza, Eusko Jaurlaritza, Euskal Herriko Unibertsitatea
Openalex Percentile: Top 10%
Quantum Computing Algorithms and Architecture
2.48
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.