Maximizing Network Reliability in Large Scale Infrastructure Networks: A Heat Conduction Model Perspective
Year: 2023
Type: conference-paper
Abstract: Reliability of large scale infrastructure networks is a critical and challenging research problem, which has attracted intensive research interests in recent years. Still, most of them focus on providing protection measures for crucial nodes or optimizing the original networks topology in varying degrees leveraging linear models for measuring reliability, while nonlinear models for improving network reliability are rarely investigated. A major challenge encountered is the high computational complexity and damage to the original network structure caused by nonlinear approaches. In light of this, we propose a nonlinear model for measuring and improving network reliability, which maps an infrastructure network to a nonlinear heat conduction system for our analysis. We introduce a new evaluating indicator of network reliability based on communities irrelevance, and further propose a new Edge Addition (EA) method to maximize network reliability, which adds multiple edges during each iteration and significantly reduces computational overheads. We successfully evaluate the proposed method on artificial synthetic as well as real-world networks. Comparison with two widely used EA methods over all tested benchmarks show the non-negligible superiority of our proposal.
Authors: Beibei Li, Wei Hu, Yiwei Li
Institution: Northwestern Polytechnical University
Language: English
Subfield: Statistical and Nonlinear Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: bronze
Award: G473269127
OpenAlex ID: W4386512337