kth.sePublications KTH
Operational message
There are currently operational disruptions. Troubleshooting is in progress.
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Adversarial Attacks on Multi-Agent Reinforcement Learning Systems
KTH, School of Electrical Engineering and Computer Science (EECS).
KTH, School of Electrical Engineering and Computer Science (EECS).
2025 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

As multi-agent reinforcement learning (MARL) is increasingly applied in real world applications, ensuring the robustness against adversarial threats becomes essential. This thesis investigates the vulnerability of cooperative MARL systems to adversarial attacks. This was achieved by implementing two types of attacks, random agent behavior and observation disruptions, on agents trained with the QMIX algorithm, within the PettingZoo Pursuit environment. The study evaluates the impact of these attacks on agent performance and coordination. Additionally, it explores the potential of adversarial training as a possible defense to the attacks. The results indicate that both attacks significantly degrade performance when applied at a system trained under standard conditions. However, agents exposed to attacks during training demonstrated improved performance during evaluation with attacks, while they performed worse under standard conditions. This suggests that adversarial training improves performance under attacks, but that there is a trade-off between performance and robustness.

Abstract [sv]

I takt med att multi-agent reinforcement learning (MARL) i allt större utsträckning tillämpas i verkliga applikationer blir det avgörande att säkerställa deras robusthet mot adversiella hot. Denna uppsats undersöker sårbarheten hos samarbetsbaserade MARL-system för adversiella attacker. Detta görs genom implementering av två typer av attacker,  slumpmässigt agentbeteende och observationsstörningar, hos agenter tränade med QMIX algoritm i PettingZoos miljö Pursuit. Studien utvärderar hur dessa attacker påverkar agenternas prestanda och koordination. Vidare undersöks potentialen för adversiell träning som ett möjligt försvar mot attackerna. Resultaten visar att båda attacktyperna avsevärt försämrar prestandan när de appliceras på ett system tränat under standardförhållanden. Samtidigt uppnår agenter som utsätts för attacker under träning förbättrade resultat då attacker är implementerade under utvärderingsfasen, medan de presterar sämre då inga attacker är införda. Detta tyder på att adversiell träning förbättrar prestandan under attacker, men att det existerar en trade-off mellan prestanda och robusthet.

Place, publisher, year, edition, pages
2025. , p. 481-487
Series
TRITA-EECS-EX ; 2025:147
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
URN: urn:nbn:se:kth:diva-376169OAI: oai:DiVA.org:kth-376169DiVA, id: diva2:2034536
Supervisors
Examiners
Projects
Kandidatexamensarbete i Elektroteknik 2025, EECS, KTHAvailable from: 2026-02-02 Created: 2026-02-02

Open Access in DiVA

fulltext(80627 kB)11 downloads
File information
File name FULLTEXT01.pdfFile size 80627 kBChecksum SHA-512
35ce0a386dafe4649eb99cbe0efdfed651a3c9044e3339612422234d17a7e8ec21d4fd4aa201500c3c7a8f57194994b78b3e0cfbd5319ecd49f18a5d8a7ff775
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Electrical Engineering, Electronic Engineering, Information Engineering

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 3922 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf