Decentralized Self-Replicating Agent Networks for Resilient Infrastructure

Bachelor Thesis (2026)
Author(s)

Tudor Coman (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Contributor(s)

J.A. Pouwelse – Mentor (TU Delft - Electrical Engineering, Mathematics and Computer Science)

B. Nasrulin – Mentor (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Faculty
Electrical Engineering, Mathematics and Computer Science
More Info
expand_more
Publication Year
2026
Language
English
Graduation Date
01-07-2026
Awarding Institution
Delft University of Technology
Project
CSE3000 Research Project
Programme
Computer Science and Engineering
Faculty
Electrical Engineering, Mathematics and Computer Science
Page Views
24
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

AI agents are transitioning from stateless utilities to long-running, autonomous systems that manage their own computing lifecycles. However, an operational gap exists between local lab simulations and production deployments. Current agent architectures heavily depend on centralized registries and naming services, leaving them vulnerable to environment blocking and single points of failure.

This paper presents a self-replicating agent network built on the open-source OpenClaw framework and the IPv8 peer-to-peer overlay network. By removing public domain name dependencies and using isolated state synchronization, the network reduces reliance on external infrastructure and mitigates failure points. This system acts as a file-sharing service, able to replicate itself into new server space from geographic locations that reduce download and response times between users and the network. Agents on the network share files among each other to prevent data loss in case of failure or targeted attacks.

Files

RP_Report_Final_Appendix.pdf
(pdf | 0.766 Mb)
License info not available