.png){kind=icon}
.png){kind=icon}
Return
TrustNet: Trust-Based Networking
.png){kind=icon}
PRE
AI
DOI：10.1109/MCOMSTD.2025.3635478.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper presents and evaluates a novel network paradigm called Trust-Based Networking (TrustNet), developed to address the increasing complexity and challenges of trust in modern networks. The TrustNet framework uses techniques such as belief function-based trust assessment, decentralized trust management, dynamic trust updates, and mechanisms for trustworthy communication between nodes to enhance AI-driven path analysis, routing, and network management. TrustNet is also designed to support large-scale scalability through cluster-based parallelism. The experimental setup, implemented in the Graphical Network Simulator 3 (GNS3) emulation environment, includes Open vSwitches (OVSs), Virtual Machines (VMs), and realistic router images. The evaluation focuses on operational metrics critical to trust-driven network services, such as network transmission time, overall convergence time, belief function execution time, and trust verification latency across varying cluster sizes. Experimental results demonstrate TrustNet's ability to maintain high levels of trustworthiness and operational efficiency in the presence of conflicting or inconsistent trust evidence. These findings highlight TrustNet's potential to redefine networking paradigms by integrating trust as a core component of network management, ensuring robust and adaptive networks.
Keywords:
Artificial intelligence
Routing
Evidence theory
Security
Scalability
Resilience
Real-time systems
Optimization
Communication standards
Data centers