Autonomous Distributed Spectrum Clustering for Dynamic User Density in Multi-Hop Wireless Networks

• Published in: Proceedings - International Symposium on Autonomous Decentralized Systems pp. 29 - 36
• Main Authors: Homma, Yuuki, Takano, Chisa, Aida, Masaki
• Format: Conference Proceeding
• Language: English
• Published: IEEE 21.07.2025
• Subjects: autonomous distributed clustering; Clustering methods; Laplace equations; Laplacian; logical network; multi-hop wireless network; Network topology; Robustness; Routing; Scalability; Simulation; spectral clustering; Spread spectrum communication; Stability analysis; user density; Wireless networks
• ISSN: 2640-7485
• DOI: 10.1109/ISADS66912.2025.00009

In multi-hop wireless networks composed of mobile nodes, such as Mobile Ad Hoc Networks (MANETs), both the mobility of communication destinations and the resulting dynamic network topology present challenges for implementing effective and scalable route control. Hierarchical routing is essential for achieving scalability in large-scale networks. However, due to the autonomous and distributed nature of multi-hop wireless networks, there is no centralized control, requiring each node to achieve hierarchical routing based solely on local information independently. Autonomous distributed spectral clustering has been proposed to address this challenge by introducing a hierarchical structure through spectral methods that utilize only local node information. While prior approaches have shown effectiveness under uniform conditions, they face limitations in adapting to increased user density. This paper proposes enhanced methods using normalized Laplacian matrices and logical networks to maintain stable cluster structures under varying node density, demonstrating their effectiveness through simulation.