A Seismic Resistant Design Algorithm for Laying and Shielding of Optical Fiber Cables

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal

6 Scopus Citations
View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)3060-3074
Journal / PublicationJournal of Lightwave Technology
Volume35
Issue number14
Online published4 Jun 2017
Publication statusPublished - 15 Jul 2017

Link(s)

Abstract

This paper considers a long-haul optical fiber cable, connecting two points on the Earth's surface that passes through earthquake-prone or other sensitive areas. Different segments of the cable are characterized by different protection levels, where a higher level through shielding represents a more costly and more resilient segment. This leads to a multi-objective optimization problem, where the two objectives are 1) the total cost of the cable and 2) the total number of potential repairs along the cable likely to be caused by earthquakes. As a measure of seismic risk, we use the concept of cable repair rate used in the civil engineering community. In our models, ground motion intensity data are used to estimate the cable repair rate, and a graph of a triangulated irregular network is used to represent the Earth's surface. We formulate this optimization problem as a multi-objective shortest path problem and solve it by a variant of the label setting algorithm. Two approximate algorithms, an interval-partition-based label-setting algorithm and an evolutionary algorithm, are also presented as methods of computational cost reduction for large-scale cases, and their results are compared. The solution leads to a Pareto front or an approximate Pareto front that enables us to choose the path and protection of the cable to either minimize cost for a given risk level or minimize risk for a given budget.

Research Area(s)

  • Cost effectiveness, multi-objective optimization, optical fiber cables, path optimization, seismic resilience

Download Statistics

No data available