SCALE DECENTRALIZED MESH NETWORK
Галина Григорівна Киричек, О. Д. Пестов, Mariia Yu. Tiahunova · Scientific notes of Taurida National V I Vernadsky University Series Technical Sciences · 2025
The paper investigates the scalability of the experimental Yggdrasil routing scheme in terms of logic, hop limit, CPU usage, and memory.Experiments with the routing daemon were conducted on different sized topologies using the meshnet-lab and coreemu-lab environments.Certain features of using this routing scheme as the basis of a large-scale decentralized mesh network have been identified.The purpose of the work is to conduct research, determine the feasibility of using the Yggdrasil routing scheme as the basis of a large-scale decentralized mesh network.The object of the research is the process of designing a largescale decentralized mesh network based on the Yggdrasil routing scheme.The subject is models, methods and software tools for implementing a large-scale decentralized mesh network based on Yggdrasil.The research proposes new approaches to deploying a large-scale decentralized mesh network as an alternative to hierarchical ISP networks.The authors identify increased CPU usage on long branches of the spanning tree and traffic anomalies introduced by a wandering root node, and propose predictive mining of small keys for deterministic placement of root nodes.They also present a system model and justify the use of the IEEE 802.11s standard with routing disabled as the basic link-layer technology.They also obtain an approximate structure of a large-scale decentralized mesh network and technical requirements for its nodes.The results show that Yggdrasil significantly outperforms other mesh routing protocols such as OLSR and Babel in limiting hops by determining system resource usage parameters.The experiments demonstrate the typical convergence time, as well as the memory and CPU usage trends of Yggdrasil and confirm that they scale with network depth rather than with its overall size.Therefore, Yggdrasil is a promising method for deploying large-scale decentralized mesh networks of independent nodes and an alternative to centralized hierarchical networks.